#619 – Super Tecmo Bug

Chris Gammell: This is The Amp Hour Podcast. Release February 13th, 2023. Episode 619. Super Tecmo Bug.

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

Chris Gammell: And I'm Chris Gammell of Contextual Electronics. Are you ready to football? Who cares about football? We're recording this during the Super Bowl. I don't think anyone cares.

Dave Jones: I thought you were kind of into that. No?

Chris Gammell: No. I'm into Premier League. I'm into... I played soccer. American soccer.

Dave Jones: Oh, I thought you played...

Chris Gammell: No. Have you seen my body type? I would get literally crunched.

Dave Jones: Right. Yeah.

Chris Gammell: I literally snapped like a stick. Yeah. I played soccer. I played soccer. Okay.

Dave Jones: All right.

Chris Gammell: Yeah. That is my preferred sport. That's any of the sports that I actively watch. Right. I will watch Premier League highlights and that's about it. Right.

Speaker ?: Yeah.

Dave Jones: I hope I don't get sucked into the soccer coaching again this week. Oh, yeah. I forgot you do that. It's tryouts. Yeah. Yeah.

Chris Gammell: Oh, it's tryouts. Well, that's a big deal. Okay.

Dave Jones: So, yeah. Yeah. And at the end of the tryouts when they sort out the teams, that's whenever they grab all the parents around and they go, who wants to be a coach? Anyone. Please. And it's like, I can remember the first time I did it. It's like, you know, we're all standing there and everyone's head's just going to look down and then, you know, and then I can see all the eyes darting towards me because I was the only one who looked like, you know, fit and he could like play and like do some stuff. Yeah. Right. With the, you know, everyone else is like, and so all the eyes and I just went, it's me, isn't it? And everyone went, yeah. Oh, man. So. Yeah.

Chris Gammell: You just got to keep your mouth shut next time. Yeah. Well, yeah. It was, yeah.

Dave Jones: Well, it was obvious. Like it looked like all the others couldn't, you know, actually run around.

Chris Gammell: You know, maybe that's a, maybe that's a second or third career here, you know, first career electronic, second career podcasting slash video, I guess. And a third career professional soccer coach.

Dave Jones: Yeah. Well, I guess the AI is going to take my job. So, you know. Yeah.

Chris Gammell: I mean, right. Yeah, exactly.

Dave Jones: Oh, boy. Oh, I will link it in actually just before this. I watched Thunderfoot's video on AI. He just released a video because he's a fellow debunker YouTuber. Sure, sure. For those who don't know him. Yeah. And does science. He's a physicist, does, you know. He's a chemist and does lots of science, cool science videos and stuff. Anyway, yeah, his video is really good on the topic. And he said, yeah, it's basically the AI, the chat GPT will dominate the YouTube attention thing. It already is. I've actually done a video on this because it's all about attention span. And, yep, it's good enough to get people's attention. You know, there's all these channels that are all generated by AI now. It's like, you know. And they get hundreds of thousands of views per video. Some even get millions of views per video. And it's like.

Chris Gammell: Yeah, but Dave, do they have that feeling in their heart that they've done something good with their day? No, they're just cold.

Dave Jones: Yeah. Cold, hard silicon. You know.

Chris Gammell: You know, I'm packing forth on the usefulness of this. At least it's the current generation of stuff. You know, some days I'm like, oh my God, I can't believe this. And other days I'm like. So like last night I was watching a movie and I was like, you know, this is a good. So I was watching Dunkirk. I still have not watched Dunkirk. Right. And it was like a World War II movie about troops. And I was just kind of curious about like percentages of like populations that had died during wars. Right. Kind of a gruesome topic. But I just was like, oh, that's a good like chat GPT thing. But it's just like the way that it spit it out was just like so. I don't know. Like so. Yeah. Yeah, exactly. And I'm sure it's gonna get better. But I just. All right. I'm sure it's gonna be a big deal a lot of places. But I'm. I think my. My interest has like ebbed and flowed quite a bit. You know, obviously.

Dave Jones: Right. Interested. Well, I am actually paying for it now. I signed up to the. Oh, you are. Oh, interesting. Yeah. Yeah. Signed up to the paid chat GPT thing. So I really. So I want to get into using it more. So I basically you pay per query. Now. And this is how it's going to work. Right. It's going to be like it is not going to be. Well, you can do it. Bing. Um, because Microsoft just had the huge announcement. Right. Where they're going to integrate chat GPT into. Bing, of course, because. Essentially, Microsoft own open AI. Now they poured $10 billion in. They got to own 49% of the company. After they've paid back the $10 billion investment, et cetera. And yeah, so they're going to integrate all that now. So we're going to have conversational search. Which is going to be absolute killer. And then did you see that Google. Google put out their response. They had.

Chris Gammell: Yeah. Yeah. Yeah. Yeah.

Dave Jones: They had their demo day or whatever it was. I don't know what it's called, you know, but. And, and, and it didn't work. Basically it, it failed. Um, and, and their share. And their market cap dropped a hundred billion dollars. Because their, because their demo. Percentage. Oh yeah.

Chris Gammell: Right. Yeah. Yeah. Yeah.

Dave Jones: But still. Right. But still. Sure. Sure. Sure. It dropped a hundred billion dollars because they failed at this demo. It's like, wow.

Chris Gammell: Yeah. Really? That is. Yeah. That's high pressure, high pressure demos. Yeah. I get some demos coming up at, uh, embedded world. And I don't think they're going to have that much impact. I hope they do. But I don't think, I don't think a hundred dollars, a hundred billion dollars with them impact.

Dave Jones: Well, you can ask chat GPT. How can I have a greater impact? Yeah. Right.

Dave Jones: Sparkling demos. Right. Yep. Yeah. Anyway. Ah, there you go. Nah, enough about AI. Rubbish.

Chris Gammell: Sure. Sure. Yeah. Nah. I have been laser cutting like a fool before these demos. Uh, and I continue to be amazed at the like availability. I don't know. Have you ever done like laser cut services? I think I've mentioned it on the show before, at least.

Dave Jones: Last time I was doing that. No. No. Well, not. It's not on my own. No. Okay.

Chris Gammell: Okay. It is so insanely easy. Like, well, I'll caveat that with, it is so insanely easy if you've used like the vector art program before. So like. Oh yes, exactly. Or, uh, what is the, the dot AI file ones? Sorry. We wouldn't say AI anymore. Oh, that's the, uh, that's the Adobe. Illustrator.

Dave Jones: Adobe Illustrator. That's it.

Chris Gammell: Yeah. Yeah. Yeah. Yeah. So like, and like, I think there's other vector programs too, but like, once you do that, you just like literally like, you just put different colors. For what you want it to be in like, you know, either like an edge cut or, uh, uh, uh, like an etching or a line line cut. It was just like so simple. It's really.

Dave Jones: Right.

Chris Gammell: I know I'm like 15 years, 15 years behind the, uh, behind the ball here, but, uh, just like the availability.

Dave Jones: Is this a local maker or is this a China?

Chris Gammell: This is actually us based, but it's Pinoco. Um, so, and Pinoco has been around for a while.

Dave Jones: Yeah. I don't think we have that here. I'm not sure what our local, I'm sure we've probably got a local.

Chris Gammell: I mean, you could ship it from anywhere. And like, and like, like all of the PCB houses are doing it too. Right. They all are doing 3d printing now. And like, that's so cheap too. It's crazy cheap. I was like doing, I was on one of the, the Chinese PCB houses and just like, you know, you put in like a model that you have and it's just, it's insane how fast and how cheap it is to get stuff 3d printed to the point where it's like, why, why do I even have a 3d printer? I know I have a 3d printer. Yeah.

Dave Jones: Yeah. Yeah. It's almost why even have your own gear, you know, it's just, I mean, it's even, well,

Chris Gammell: how about this? I, so I just also got some, I'm in demo season, right? So I'm building all these different demos and I'm getting like these custom cut cases, right? So off the shelf cases. And I talked to the manufacturer and I'm like, Hey, can you like, I sent him a drawing and I'm like, can you cut out this part of the case? Cause I want to put something in there. No, yeah, that's fine. And then, um, and they, you know, I asked them for the STL of like the modified drawing that they did and they're like, nah, I, they're like, no, that's proprietary. I can't send that to you. Oh, okay. Well, why? I'm thinking I didn't respond. I'm like, yeah, that's fine. It's your business. That's fine. Whatever. I was like, why would they do that? And then literally like two days later, I'm like on this site and, and I'm trying to figure out like, oh, well, I wonder how much that would the cost to like do this, um, to like get a case printed, to like print my own case. And I was like, oh, actually I have a 3d printed case model from, from off the website. I could just use that. You could just, I'm like, oh, oh, actually that's their design. Yeah. I shouldn't, I shouldn't use that. But it's like that, that right there was the explanation of like why they, it's, it's,

Dave Jones: although it's not going to come out the same. Cause we're talking about a, uh, you know, a stamped case as opposed to like an injection molded case versus a, you know, 3d one, but it's getting there and like, and then

Chris Gammell: you look at all the finishing options too. And like, you know, you can just, I wish I was better at 3d modeling, honestly, because if, if you're good at 3d modeling these days, like we talked about with like the prompt engineering, right. That kind of thing. If you're just good at like modeling and a shape, it's, it is amazing what you can get made these days. Like just with the click of a button, it's literally unbelievable. It's, you know, I gave a talk. I don't think it actually was re-released through, um, Hackaday yet, but at Supercon this year, I gave a talk about the one engineer dev shop and it was kind of based around this idea of just like the availability of services and how much you could pump out without needing to have all of the people you need to have in like a vertical engineering company in the past.

Dave Jones: Exactly.

Chris Gammell: And it's not like you couldn't scale it the same way. You couldn't scale it to the same like production way levels, but like that prototype, you can do so much with that prototype versus what you've been able to do in the past.

Dave Jones: Yeah. It's absolutely incredible.

Chris Gammell: Yeah. All I got to do is dream. And then, you know, chat GPT will write the, uh, and chat GPT will write the, the copy then at the end.

Dave Jones: Well, good stuff. Oh boy.

Chris Gammell: Yeah. Yep. Yep.

Dave Jones: I've been dealing with, um, uh, bugs.

Chris Gammell: Yeah. Like real bugs?

Dave Jones: Yeah.

Chris Gammell: I'll show you. Those are, those are formidable.

Dave Jones: Right. You eat Z bugs. You will eat Z bugs. You will buy your $20,000 oscilloscope and eat Z bugs. Yeah. Anyway, I've been playing with the latest, uh, Roden Schwartz, um, um, MX04 scope. I've been doing a video. I'm editing the video right now and it's got, it's long. Um, but yeah, I've been, um, I've been, um, on a, uh, like a bug hunting session for them. Um, it's just. Really? Yeah. Yeah. Yeah. What's the old behavior? It's doing all sorts of stuff. Um, yeah. The, the latest one is it actually locks up just using the horizontal control. So all I'm doing is the horizontal control and I can get it to lock. I can put it, I can get it into a mode where I can lock it up and it's reproducible. And it's like, you know, so I made a video of it showing them and they go, holy shit. And this is in addition to one that I did a couple of months back where I also got it to lock up in a UART decoding mode. And I think, you know, it was a pretty obscure thing that I was doing. And it turns out, I might've even mentioned on here, it was the board rate changing the board rate too quickly caused it to lock up, caused the sampling subsystem to lock up. And it's like, you know, I thought, okay, that's a pretty obscure bug. And now I got it to do it again, just using the horizontal sweep control. And it's like, geez, I'm good. It's like, you know, I feel kind of chuffed when I actually find these bugs, you know? And, but it leads me to think that, I mean, these scopes are so ridiculously impressive in terms of the capabilities that they have. Like, I don't know how many lines of code would be in this thing, but it's insane.

Chris Gammell: As much as that's a good measure too, right?

Dave Jones: I mean, it's, I just, it's one way to benchmark for sure.

Chris Gammell: Yeah. Yeah. Yeah. But I'm expecting that. They run pretty lean actually, which is surprising. Oh yeah.

Dave Jones: Yeah.

Chris Gammell: You know, so they would, they might have like a common chipset, right? That was designed five years ago that is like bleeding edge, of course. But then you think about like the embedded teams, those are not, you really think about the embedded and then all the software on top of it as well. And it's just like, those teams, I think get stretched pretty thin actually. When you think about like the bugs that might be popping out, the bugs aren't, probably are not in the silicon. They're probably in the computing and like the UI and things like that, right?

Dave Jones: Well, I think it's in the VHDL. I think it's in the, although, no, this one has a custom ASIC. This one has a new custom.

Chris Gammell: I think most of them are custom ASICs, right?

Dave Jones: No, no. Most, most modern scopes are, use a Xilinx, what you call it? Zinc. A Xilinx Zinc. I didn't know that. Yeah. Yeah. Most of them do. It's only.

Chris Gammell: How would you know, Dave? What do you ever open these things up or something? Geez.

Dave Jones: Need scopes up. Yeah. Yeah. Just occasionally. Yeah. This one uses a new, you know, super whiz bang. In fact, it's, I believe inside it's the ASIC and then it's the processor and two ADC front end chips. And that's pretty much it. You know, it's like everything's in the ASIC.

Chris Gammell: What is the ASIC in that case? So the ASIC is doing like.

Dave Jones: The ASIC is the Samstream subsystem. It's doing the FPGA. It's doing, sorry, it's doing the FFT stuff because this has insanely fast FFT, like, you know, 50,000 per second. And you can do it with all of the stuff turned on. You know, it's just.

Chris Gammell: So like the horsepower kind of stuff. Yeah, it's the horsepower.

Dave Jones: It's a four and a half million waveform updates per second, which we're at now, right? Which is four and a half times quicker than the Keysight one, which used to be, which is a million waveform updates per second. I thought that was impressive. This one's four and a half million.

Chris Gammell: I have no need for this at all. Like this is. There are totally people out there that need this. I know.

Dave Jones: You need this in very niche, you know, circumstances. It's like you don't need it for an everyday use scope. You're just probing some signals, right? Although the FFT mode in this is just beautiful. Like it's, you know.

Chris Gammell: That was a chef's kiss, right? That's what that was. That was a chef's kiss.

Dave Jones: That was. Yeah. Like it's like a real spectrum analyzer as opposed, you know, if you've used the FFT function in your scope, you go, oh, this is not a real spectrum analyzer. It's just like, eh. You know, they're kind of mediocre. You know, kind of, you know, you can see. How many buckets do you get, right? And that's what it really comes down to. It's like, yeah, no. But this one looks and feels like a real spectrum analyzer, right? It's absolutely fantastic. And it was doing, you know, 50,000 FFTs per second. And, you know, it's just nuts. And it's got, yeah, but it's all like, I suspect it's in the ASIC, but I don't know how, like if they've got like state machines inside the ASIC, like you can have an ASIC and then you can run state machines inside there. It depends how you design it.

Chris Gammell: Wait, so if you're triggering this though from the UI, what is, I just feel like a lockup is always at UI levels, like in the, you know, some kind of graphics engine, something like that.

Dave Jones: Possibly. But see, here's one of the weird things. Here's another, I think it's a bug, but it's like, it's more of like a system architecture thing, right? As I said, I can get four and a half, I've measured four and a half million waveform updates per second on this scope, right? And then I start turning on, you know, measurements. It's still, you know, it's still fine. I can turn on, you know, cursors. It's still, it's still the same update rate. It doesn't actually reduce it, right? Cursors.

Chris Gammell: Why would cursors do anything?

Dave Jones: Because they, because they take computational time.

Chris Gammell: The subsystem that's doing, yeah, but they're totally, no, come on, man. That's totally different systems. The UI has got to be completely separate from anything that's doing like hard math.

Dave Jones: Have you used a modern scope? No. Right? The Keysight, yes, because Keysight have the custom ASIC and serial decoding and FFTs and all sorts of stuff is done specifically in the ASIC hardware. So you turn on all the measurements, doesn't slow it down. But in this case, right? I turned on, right? So I turn on all my vertical measurements, didn't slow it down. I turn on one horizontal measurement. Frequency measurement, right? Boom. It drops from four, from a million waveform updates per second to, to under a thousand. By turning on that one thing. And I go, what the heck's going on here?

Chris Gammell: Somebody forgot to push that into the fabric, I guess, huh?

Dave Jones: Like, yeah, there's something in the architecture there that was, that, you know, causes like three orders of magnitude drop in update in waveform processing and update rate speed. Just because I turned on a horizontal measurement. And, you know, so they're checking. It's crazy too. They're looking into it. And it's just.

Chris Gammell: You think about like all the, all the, like the functional testing you'd have to do. Like every time you push a change to that kind of software and like, there's all like the UI testing and I'm sure all the code, you know, code checking and linting and all that happens there. But then you just think about like, yeah, there's just the use cases of like, I don't know if they automate like knob turns. They must have some kind of tester that can do that sort of thing. But like just putting it on a, not a test stand, but almost like an, like an audit, an audit test, right? You're just saying like, okay, we had to run through every single test with every single configuration, you know, how much time you can.

Dave Jones: Well, you can't. Like you can't run it through every test, every permutation. You can't, you physically can't do it.

Chris Gammell: I mean, you could, but it would just take a long time, right?

Dave Jones: No, no, I, no, I don't believe you could. I believe the permutations would be the same number as like chess or something like that.

Chris Gammell: Dave doesn't care about infinity. That's what we're hearing.

Dave Jones: It's insane. I can, I can come up with scenarios you wouldn't even think of. And, you know, and then other people come up with scenarios that I can't even think of. Right. No, it's, it's just absolutely nuts. Right. Who would have thought that turning the board rate control too quickly, too quickly would have caused a complete system lockup, right? That would be hard to test with a, uh. Who would have even thought to test for that?

Chris Gammell: You wouldn't, you wouldn't probably have a test for, tester running when you're switching between them and nor would you make that variable. Yeah, that's a good point.

Dave Jones: Exactly. And there's, and there's tons of permutations when you go down the, you know, turtles all the way down, you know.

Chris Gammell: Turtles. Yeah. Turtles everywhere.

Dave Jones: Anyway. Yeah. So there was some discussion on the EV blog forum about this and somebody said, oh, I can't, you know, I showed the video of it, you know, the scope locking up and everything. And, oh, this is ridiculous. You know, you pay $20,000 for a scope. This is unacceptable. And, you know, this had never happened if this was a car, for example. Would you buy a, you know, a $30,000 car and you expect it to just work? And it's like, yeah. It's like, have you ever driven a Tesla? Tesla, they're infamous for like, you just being like software, you know, bugs and things locking up and doing all sorts of stuff. Right. And, and it's like, no, you can't make that, you know, once the system gets so complex, it's like, yeah.

Chris Gammell: Yeah.

Dave Jones: Is it inevitable that you're going to have four?

Chris Gammell: Don't need it.

Dave Jones: I don't know. There's, there's, I'm sure there's system engineers that get, oh, when I was a boy, you know. And, um, um, a guest, which, um, hopefully we're going to have, um, soon on the, uh, podcast. Um, he mentioned on that same forum thread that, oh, this is because we, uh, nobody thinks like an ASIC designer anymore. Cause everything's FPGA. They can change it. They can push field updates and stuff like that. They don't have the same design mentality of, you know, absolutely making sure the architecture and the testing is a hundred percent, you know, or as close to a hundred percent as you can get it.

Chris Gammell: The fear of God. Like being counted in view of the masks that being on.

Dave Jones: Like I'm spending a million dollars on this ASIC. Like I'm spending 10, he, he, he mentioned he's worked on ASICs, which are $10 million, right? For a spin. And it's just, you know, I'm sure we'll hear about this on an upcoming episode when we get him on. But, um, it's just, yeah. Like, you know, he, he mentioned that, yeah, the ASIC designers have a different viewpoint. Cause they know there's so much on the line. Whereas if you know, oh, I can just, you know, push out a firmware update later. I can push out an FPGA update later. You know, we can, we can just fix anything in the field these days. Does it lead to a less strict.

Chris Gammell: Different mindset.

Dave Jones: Yeah. A different mindset, mindset, more, more lapsed rule, you know, rules, less testing. I think it probably does. Right. And is that good or bad? I don't know. It gets a product to market quicker. It's like, yeah.

Chris Gammell: I mean, there is a cost. There's like a time cost to actually do that, to do that testing and crazy validation and verification and stuff.

Dave Jones: And then how do you even know? Right. So.

Chris Gammell: Yeah. I mean, this is, um, past guests, uh, Claire Wolf talked about this a little bit about like, uh, you know, like even if you did like completely random inputs, that's not. Enough stuff or that's not always a realistic thing. So you have to do like some amount of random, but then it's a subset of all random possibilities and stuff like that. And yeah. It's nuts.

Dave Jones: And then do you, uh, can you, uh, can you architect your stuff so that you can guarantee that? Okay. This, you know, subsection, you know, if I turn this horizontal control over here too, uh, fast, it's no, there's no way it can lock up the sampling subsystem over here. Cause they're, you know, they're totally separate. And from an architecture point of view, it's just not physically possible. So you don't have to test that, you know, is it possible to engineer?

Chris Gammell: I'm not even sure how those are possible right now, but I don't know. I don't know. I don't know if about scope. So I'm just going to, I'm going to bow out on that one.

Dave Jones: And then, and then like, I just did a recent video on the, uh, tech, the tech two series, you know, that sexy little portable, uh, one that they've got the new design.

Chris Gammell: Um, uh, Nope. Didn't see that one.

Dave Jones: Nope. Thank you. Anyway, the tech series two tech series two. Very cute. They released this new form factor. It's slim, almost tablet-like. So it's sort of like a combination of a tablet bench scope all in one kind of thing. Anyway, they actually released that to market, went out. They've only, what's been six or nine months now, and they've only finally gotten around to enabling the pattern generator on it. So they sent me the pattern generator actual update, right? So they sent me the pattern generator update. And so I installed that and I actually did a video. It's on my second channel. You can watch it if you want. And, uh, so I installed that and I was playing around with the pattern generator. It's really cool and everything, but sure enough, found a bug, right? And so, uh, what was the bug on that? Um, it was, that's right. It at, at high board rate settings, right? Like everything worked fine. Like I was generating a pattern from a CSV file, right? So I didn't, like there was an Excel file that had, uh, data in it. And then I was using that to drive the pattern generator and then the pattern generator, and then I was simulating a, uh, UART and that worked fine at low speed, right? It worked fine at 9,600 board. It even worked fine at one meg board. But when I took it up to 10 meg board or meg bits per second, it, it didn't, it didn't work anymore. And I'm going, what, what the heck's going on? And I finally realized it died at about eight meg or something like seven or eight meg. And then I, uh, talked to tech and they went, oh yeah, oops. It looks like the, uh, the, uh, pattern generator bit rate was not matching what we said it was or something like that. So they fixed it, bam. But right. This is a release. They re they released the new firmware. Here's this pattern generator. You know, it's an option that costs, you know, 500 bucks or a thousand bucks. I don't know. Um, and boom.

Chris Gammell: For the people who need it, they, they really want it. Right.

Dave Jones: Yeah. And yeah. Well, for people who need it. And I just found a bug straight away and they just, and it was like, it was like something simple. Well, you know, simple in quote marks, right? Like the bit rate, the actual output bit rate, I believe didn't match what it told you on the screen or, or what you actually said. And it was like, I don't know, maybe they got a multiple wrong or something like that and worked at low frequencies because there was enough margin. Higher frequencies. Nope. Nope. Completely.

Chris Gammell: It's interesting that bug, bug bounties have never made their way into the hardware space ever.

Dave Jones: No. Have they, have they, has anyone got any, an example?

Chris Gammell: I've never, never heard of one. I've never heard of one other than, I think because it's like probably the, the distribution is not wide enough.

Dave Jones: Right.

Chris Gammell: Or you think about like, like a bug bounty on like an iPhone, like that is hugely impactful across the world versus like, I mean, how many scopes that you're reviewing?

Dave Jones: Is this a software and hardware nerd thing? Cause like do hardware nerds go, oh, I couldn't be bothered. I just want to use my scope. I don't want to systematically test for, you know, bugs and stuff.

Chris Gammell: I think there's, there's probably not enough volume to make it worthwhile for someone that can do it full time. Yeah.

Dave Jones: Cause there's only X, X percent of people are going to be willing to put in the effort to do that sort of thing. And if you're only selling, you know, 10,000 scopes as opposed to, you know, 10 million iPhones, then. Yep. Right.

Chris Gammell: It's probably cheaper for them to just hire like integration testers in house that run through this whenever they do a release or whatever, or make the FEs do it or something like that.

Dave Jones: Yep. And anyway, what I'm getting at is that like, it's not one individual manufacturer that's bad at this. Like all of them right, right across the board.

Chris Gammell: Like, you know, it's hard to do.

Dave Jones: Especially when they get more and more complicated. It's just, yeah. Right. Right.

Chris Gammell: And, you know, faster release times they want to, you know, I think I'm sure every industry gets pushed by, you know, managers looking at it, like being like, well, they release a new iPhone every year or whatever. And it's like, yeah, that's a different, different echelon of development and then cost and everything like that.

Dave Jones: Well, the, the latest, uh, Bryman multimeter that I sell, the BM, uh, seven, eight, six, it, uh, they've moved into shock horror. Um, they've, they've stepped into the 19, late eighties, early nineties. Um, and they've decided, no, we're going to put a flash firmware micro in our meter. Right. So that we can actually do bug fix updates. Whereas before, uh, before, or most multimeters on the market have a one-time programmable chip in there. So the, once, once you buy that multimeter, you can't like, yeah, maybe you can get them to send you a new chip and you can de-solder it, you know, solder it in a new surface-bound chip. Right. Um, which they have done, um, for, for certain people like myself and other YouTubers who, you know, do that. They, they will do that for Joe blogs, uh, customer. No, they're not going to send you. It's like the, the, the multimeter you buy is the multimeter you are stuck with. Right. Cause it's, it doesn't have a flash programmable thing, you know?

Chris Gammell: So, yeah, actually that's a good, uh, that's a good segue there. So, uh, past guests to the show, uh, uh, Ooh, Jay Carlson, little, little brain blip there. Uh, Jay, uh, released a, uh, a blog post recently and Jay's had some pretty prolific blog posts, but this was actually about the chip we've talked about on the show, the, uh, the RV 32 10 cent risk five microcontroller. And basically he did an analysis. He's like, you know what, this is actually not the best chip you can get for cheap. And, uh, and here's why. And, uh, and so instead I forget what the, uh, what the chosen part was, but it's basically, it's a, it's a cortex M zero, um, that is on LCSE right now. Um, and it's basically 48 megahertz as well. It's made by a memory manufacturer. Um, post this link to Dave. So he has it as well in front of him. Uh, good analysis. I mean, like usual, I think the only thing so like, and you know, I think Jay wrote it kind of in an express exasperated manner of like, why do people care about this other thing when there's existing parts, uh, fit, uh, this, the, the guidelines, the POUYA PY 32, whatever. Um, some of it is just the, uh, the fact that it's risk five. Like that's, I think that's some of the interest from my perspective, at least. So the CH32.

Dave Jones: It's also fairly powerful at 10 cents, but it's kind of novel that it's a risk five, but you know, and there are people who assume that because it's a risk five, the reason it's so cheap is because you don't have to pay the, um, royalty on it. So anyway, so how much is this, how much is this cortex jobby? How much is this?

Chris Gammell: A sense up to 74, depending on the, uh, configuration and quantities. Yeah. And there is a ton of different things. I think one thing that he points out that is a very good point is the, the different, uh, uh, packages that they're smaller than what's available. Otherwise, I haven't seen on the RV 32. I hadn't seen anything that's like, so like the, the boards that you and I got, the test boards are like TSOP twenties, right? So very, pretty chunky matches kind of like the STM eight and, and similar, like, uh, in similar parts that are out in the marketplace. And, uh, so they're, they're trying to replace that sort of stuff. Yeah.

Dave Jones: Well, I'm, I'm looking at the cheapest one. The only one that's 10 cents or cheaper is the eight, either the eight bit or 10 bit, um, the eight, sorry, it's either the eight pin or 10 pin. Uh, package. If you go higher than that, you start getting into.

Chris Gammell: Right. Right. Yeah. So then that does limit you on GPIO and stuff like that. And so, um, yeah. And you know, Jay knows the stuff too. So like if you, if people don't remember Jay has been on the show, but also talked to the, his articles about, uh, the two, two kind of like blockbuster articles, one about the amazing $1 microcontroller. The other one being, uh, the, uh, where he was building Linux systems. Um, so we'll link both of those in here as well.

Dave Jones: Right.

Chris Gammell: Um, yeah.

Dave Jones: So, yeah. So, so it's, so it's the poo. The P-Y-U. Poo. Poo is, is the manufacturer. P-U, sorry. P-U-Y is the manufacturer.

Chris Gammell: P-U-Y. Oh yeah.

Dave Jones: P-I-I. I'm staring at the data sheet now. Oh, it's an A at the end. Okay.

Chris Gammell: Yeah. Yeah.

Dave Jones: P-U-Y.

Chris Gammell: P-U-Y.

Dave Jones: P-U-Y.

Speaker ?: P-U-Y.

Dave Jones: P-U-Y.

Dave Jones: P-U-Y-O-O-2 for those playing along at home. And the data sheet is in, uh, Mandarin. Unfortunately.

Chris Gammell: Yep. Yep.

Dave Jones: Yep.

Chris Gammell: Yeah. I think the, I think the other thing that does potentially work in this part's favor is if it is in the, you know, arm, uh, Cortex M zero plus family, then you're going to just get more tools there as well. Right. So like, so Dave did a show, a video about the, about the moon river and like, it is possible to use that, but that is the only tool chain that I've seen so far. That is M-O-U-N. Yeah. Moan. That's how you spell moan. Moan is M-O-A-N.

Dave Jones: Oh, okay.

Chris Gammell: Right.

Dave Jones: Is it? I don't know. I can't spell it.

Chris Gammell: I would still say M-O-U-N would be moon to me. Like you.

Dave Jones: Yeah. Okay. Yeah.

Chris Gammell: Uh, yeah. So. Yeah, it's good. I mean, it's good. It's a, Jay's always very good at like, you know, gathering data and this sort of stuff. So great article to check out. Um, I think, you know, it's interesting just seeing that there are other parts that are out there, you know, like just the fact that like, these are all targeting, I guess, toys. I mean, like, I don't know. Like it's, it's all low end kind of simple, simple stuff. But like these things are everywhere, right? The STM 32, F zero, three, zero. Like that part is everywhere. Got it. And you know, the V, the V 100 or sorry, not the V 100. The, um, uh, F 100 family as well are like very, that's like what, uh, the GD 32s. I forget what the GD was. I just remember GD. Cause I always thought it was like goddamn 32. Uh, yeah, the, that was another one that clone that was like the pin per pin for pin and register for register clone of the STM 32 that's out there too. And like, you know, these, these companies wouldn't be doing this unless there were such a large market that they're trying to just kind of nip at the heels of ST.

Dave Jones: Oh, there's a huge market in China. It's, you know, it's absolutely massive. It's crazy.

Chris Gammell: Yeah. So.

Dave Jones: Yeah. You have to shave five cents off the cost of the chip. So, you know, yeah.

Chris Gammell: Right. Yeah. Right. I'll do whatever it takes. Right. Yeah. Switch, switch to a different, uh, Windows based, uh, IDE. No big deal. Yeah. That's worth it. Sometimes, however many devices, products I make, you know, it's crazy. Exactly.

Dave Jones: All right. I, I still haven't decided what a hundred bucks worth of micros I'm getting from. Yeah.

Chris Gammell: Maybe, maybe you could split it, uh, at risk five and SCM 32 and then you could, you know, maybe, uh, I guess you can't use a hundred dollars towards laser cutting, but, uh, you know, I was gonna say you could, you could make a whole product. You could make a whole product, uh, demo for a hundred bucks. Yeah.

Dave Jones: Yep.

Chris Gammell: That actually would be kind of an interesting challenge, right? Would it be like, what is the coolest, uh, the coolest prototype you could build for a hundred dollars? You only have to make one of like the most complex thing that you could possibly build and you have to like show receipts. So there's only a hundred dollars.

Dave Jones: Yeah. Yeah. So you've got to, yeah, you've got to get the PCB one off. You've got to get it delivered. You've got to get the, you know, the laser cut case delivered, the 3D printed thing delivered and, you know, the parts delivered and.

Chris Gammell: Yeah. That seems like it's a, there's like a contest in there waiting to happen. Yep. Yep. There is.

Dave Jones: I, I, I will say you can, you can assemble it yourself so you can solder the parts. Sure. Sure. Sure.

Chris Gammell: I think, well, but I think the other thing too, is you need to have like other rules of like, you also have to charge for like electricity. You have to charge for like filament. If you're going to do your own printing, printing, I think you really, it has to be like, yeah, no labor costs. Right. So, but definitely like, if you're assembling, you better, you better put a spreadsheet together with all the solder paste you use too. Right. I mean like to that level, I feel like you really, cause like it would really have to be optimized that way. Like maybe no, did I say electricity? Maybe, maybe no electricity. Yes, you did. Yeah.

Dave Jones: That's, I think that's going a bit far.

Chris Gammell: Yeah. Well then, so then like, could you have your own laser cutter? Do you have an advantage then too? Right.

Dave Jones: Right.

Chris Gammell: Cause like the, but like wear and tear of the, the, the tube in the laser cutter. I don't know. You'd have to have. No, no. You'd have to have some decided.

Dave Jones: You'd have to outsource everything. I reckon you've got to outsource everything. Are they like, right? Because basically everyone's got a solder and iron, right? So you can go, okay, I've got my solder and iron. I can, I, you know, you can do basic assembly yourself, but you're not allowed to make the board yourself. You're not allowed to, you know, 3d print yourself. You're not allowed to laser cut yourself.

Chris Gammell: Yeah. So it's really just about finally the cheapest. Yeah. That would probably be what would happen is you dig out a bunch of people that like had like their secret, like low cost services.

Dave Jones: Right.

Chris Gammell: Yeah. I don't know. Shipping might eat all that up though too.

Dave Jones: Oh yeah. Yeah. Totally.

Chris Gammell: Yeah. Yeah.

Dave Jones: It depends on where you are.

Chris Gammell: Yeah. That's tough. You know, maybe this is where the challenge is. Sounds like a hackaday contest. I mean, we should call up the old hackaday folks and see if they want to do that.

Dave Jones: Right.

Chris Gammell: Or, you know, we could, maybe we could publish a yearly index of like, what is the type of prototype you could build for a hundred dollars? You know, like how they do like a basket of goods when they're doing like economic measures and like inflationary stuff. So we could be like the amp hour a hundred dollar prototype index.

Dave Jones: You can do all the work in compiling that. Thank you very much. Yeah. Yeah. Yeah.

Chris Gammell: Yeah. Yeah.

Dave Jones: Yeah. Nah.

Chris Gammell: Nah. Yeah. Nah. Okay.

Dave Jones: All right. What else we got on the list this week?

Chris Gammell: Hmm. What else do we have on the list? There is another. Oh, speaking of, well, this is probably a hundred dollars by itself, but Google is selling a microcontroller development board and it's canceled. Oh, okay.

Dave Jones: I was going to say, isn't that canceled? Yeah.

Chris Gammell: Yeah. Yeah.

Dave Jones: So they've terminated the RISC-R5 development project.

Chris Gammell: No. So this is actually, they've been making a, yeah, this is a different one. They've, Google has been working on a chip TPU. So a TensorFlow processing unit. Uh, yes. And then, so basically they, this coral is that the name of that. And I had talked to someone about maybe coming on the show about it because, you know, it's interesting. And basically it's like, if you're trying to do like vocal processing and things like that, then having, having an FPGA is obviously very expensive in terms of silicon and power and programming time, whatever. And if you're just kind of making an inference engine, that's doing the same stuff over and over again, I'm talking about, but I have no idea what I'm talking about, but having that in ASIC apparently makes it a lot easier for schlubs like me to do something like this. Right. Uh, and so if you're doing like, like it has a camera on board, so if you're doing like object detection, uh, doing that in custom silicon that is specialized for detecting images and edges and stuff like that, that's how you want to, you want to operate. Uh, okay.

Dave Jones: It's an edge tensor processing unit.

Chris Gammell: Yeah.

Speaker ?: Yes.

Dave Jones: Yep.

Chris Gammell: Cool. Uh, and so, and like, and so Google, I think the coral group.

Dave Jones: Hang on.

Chris Gammell: Oh, all right. Live delivery.

Dave Jones: Sorry about that. That was the UPS guy. That was the brown chair guy. All right. Yes. And, um, the, the tracking was lacking. I must say, cause you remember just before the show, I said, Oh, I wonder, I, there was a package coming. I thought, Oh, maybe it'll have.

Chris Gammell: I've had such a crap, crap experience with, with couriers lately. Really?

Dave Jones: Okay.

Chris Gammell: Nine days to get a two day delivery to my house from, from Thief River Falls. It was very frustrating. Nine days. DigiKey.

Dave Jones: Nine days for DigiKey.

Chris Gammell: No, DigiKey handed off the thing fine. It was FedEx, FedEx problem. Oh, okay. I got an email that said like, Oh, we're having, you know, weather problems in Memphis. Say, okay, that was a day. Yeah. What about the other seven days? You know, like what, what happened guys?

Dave Jones: That screws up all the schedules. You know, if you have unexpected events like that. I don't care.

Chris Gammell: I'm looking out for number one, Dave.

Dave Jones: You might know.

Chris Gammell: One person dev shop, you know?

Dave Jones: Yep.

Chris Gammell: Stuck, stuck waiting for, waiting for FedEx. I could have ordered another set of board or another set of parts. Anywho. Yeah. I was very frustrated. FedEx is. Could you?

Dave Jones: While you're waiting for that, it'd be interesting now if you could use all of the same parts again from DigiKey and then use another courier to see if they get there before the other.

Chris Gammell: I guarantee they would have. I got, I have, how about this? So I got my laser cut parts in two day shipping from USPS from all the way across the country. So like, yeah, it would have been fine. Yeah. That would have been cool. That would have been cool. Let's just logistics. Like I, I, I am so, so grateful for the time we live in with logistics. Oh, it's crazy. And I cannot stop myself from complaining all the same. It's, uh, yeah.

Dave Jones: No, I've, I've seen massive changes over the last 30 odd years. It's just nuts.

Chris Gammell: I can only imagine. Yeah.

Dave Jones: Yeah. It's just absolutely incredible. Yep.

Chris Gammell: So, uh, this board, the coral board is, so basically coral, I believe is the internal group at Google. That's basically doing ASICs that can go into processing centers to do, you know, this kind of processing and server racks as well. But then they also decided to spin it out as a, I think this is a, I think thing. Uh, and now they also have a microcontroller version. They can talk to it. It's got an NXP, uh, RT one, one, one seven on it, which is like one of the crossover chips, which is a beast in its own, right? Cause it's like a 500 megahertz chip. Like it's insane. So like this thing is really, really beefy what it is. And I don't care. I'm not going to buy it because hell no. I don't, I don't, I don't, I barely trust Gmail these days that they're not going to just up and cancel that thing. Let alone a microcontroller. Right.

Dave Jones: So they haven't actually canceled it. You're just anticipating them canceling it.

Chris Gammell: Oh my God. Yeah. It's exactly what I expect. Yeah.

Dave Jones: Okay. Let's, uh, let's not be jaded. Let's note the date, uh, the 13th of February.

Chris Gammell: Yeah.

Dave Jones: In 2023. When will the announcement come that, uh, I mean, you've canceled the coral.

Chris Gammell: And like people are going to build their product. They're going to build businesses around this sort of thing. Yeah.

Dave Jones: Yeah. Exactly.

Chris Gammell: Yeah.

Dave Jones: So what exactly can this tenser, um, processing unit do? What exactly does it do?

Chris Gammell: What does, what does this say? What does it say you do here? Well, I have 4 trillion operating operations per second with eight into eight precision. Uh, I think what it does, Dave, is basically it, uh, it has, it does wake words. It does image recognition. It's meant to do it at lower, lower powered levels than you would do that in general computing. Right. So you can also do like TensorFlow light. So if you look at companies like edge impulse, they operate on, um, special or general, generalized processors, microcontrollers specifically. Right. And that's kind of what they specialize in. And then you can also do things like pie torch and TF light and all these other crap things that I don't understand. Uh, but basically you're trading a model and then you're putting that model into a, an engine that can run that model, either a generalized computer or a specialized computer. I'm pretty sure that the coral tensorflow light micro is the latter. That it's a specialized thing that runs it at a better, better clip and a lower power.

Dave Jones: Okay.

Chris Gammell: More better. Exactly. And then, and then there's a microcontroller in front of it as well.

Dave Jones: Got it.

Chris Gammell: But man, like they really, like, I'm really like the, the fact that they put a 500 megahertz processor in front of it, just like, that's what's crazy. I've used the RT line from NXP and they're beefy parts. They're really beefy. Yeah. Um, yeah.

Dave Jones: I'm sure I'm not understanding this stuff either, but I've been amazed by the raspberry pie. Like I've, with, with the kids, I've just, just built one of those retro pie, uh, consoles, right. Which uses a raspberry pie three. You will run, you run the retro pie operating system and it basically emulates every arcade, you know, uh, every arcade machine ever, uh, you know, made right. It didn't, you know, not only a, you know, old Atari 2600s and stuff, but it does, uh, playstations and it does, you know, all the, um, yeah. Right. It does them all. It emulates like dozens and dozens of different arcades. And it's like, it's, it's insane. Just in one little chip. How much does the raspberry pie, uh, three processor chip cost on its own? Like, well, well, if you could buy it, you cannot buy that.

Chris Gammell: You can buy that. If you handed, if you handed Eben a, a bag of cash and signed every NDA, I'm pretty sure you still couldn't buy it.

Dave Jones: Right.

Chris Gammell: It would, you know, it'd be barrage time really.

Dave Jones: Yeah, but still, what is it? Like, how much does the board sell for before the, uh, price? 35 bucks.

Chris Gammell: Nominally is 35 bucks. Right.

Dave Jones: So the chip's like 10 bucks. I mean, that's an old version too. 15 bucks, right? Yeah. Yeah. Yeah. It's an old version, but I'm just absolutely stunned. Um, and, and, but gosh darn it, we tried to play, uh, a, uh, PlayStation game on it. And apparently the PlayStation emulation just isn't tight enough. It's not, it's not written well enough to actually execute on a Raspberry Pi three. Gosh darn it. But like, it's just incredible how you can just emulate entire things. Well, I mean, with all those custom ASIC chips and they, they just emulate it. Yeah.

Chris Gammell: But some of that stuff, if you look at the, just the horsepower of like some of the original, like even original PlayStation, like it's just dwarfed by everything that you have. Like I think about the embedded computing that I use for IOT devices and like that.

Dave Jones: Yeah. But the fact that it does a real time emulation of it is just like, it's like, it's just nuts. On a general purpose computer. It's just software. What's the hard part? What's the hard part?

Chris Gammell: It's just software. Just software.

Dave Jones: Anyway. Yeah. So, so, so if you don't know, it actually is a true emulation because you download the ROMs from what you do. You download the ROMs of the game. So it's actually the ROM image. So it actually simulates the processor. So, and it simulates the, you know, it simulates everything and it just runs the original ROM. It's just nuts. So hats off to the people who've written this stuff and hats off to, you know, the evolution of processor tech. Like it's just.

Chris Gammell: How do your kids like retro games? I always wonder if like.

Dave Jones: Oh, they, they, they don't play modern ones. They do not play modern ones that look realistic. They will not play games that look realistic. So they're into the Lego games and stuff, right? Anything that's Lego. Right. So it's, it's got to look, it's got to look Minecrafty. It's got to, you know, it's got to be that pixelated look, you know? Yeah. Yeah.

Chris Gammell: But that's, that's been a, that's still like, there's like a middle ground there. I'm saying like, if you put them in front of like a ET game on the Atari.

Dave Jones: Oh, right. Okay.

Chris Gammell: Which you would not.

Dave Jones: But that's pretty bad. Yeah.

Chris Gammell: You put them in front of like Castlevania or Wolfenstein, you know? Oh, Wolfenstein.

Dave Jones: No, they, they enjoy that.

Chris Gammell: Yeah. Okay.

Dave Jones: Yeah. They've actually played it.

Chris Gammell: I guess that's not that. No, it's still old. You know, Dave, I'm having a hard time even realizing how old I am at this point. Like I know I'm old. Like I'm, I am, I am quite close to being over the hill.

Dave Jones: And I spent a week downloading Castle Wolfenstein from a BBS, from a dial in BBS. Oh yeah. Okay. Yeah.

Chris Gammell: Yeah. That's the hard part. You know, that's actually the hard part playing, playing the games in, in the Jones household is to, you know, you boot up a new game and Dave has to tell you about how he downloaded from BBS.

Dave Jones: Yeah. Yeah. Yeah. Yeah. Back in the day.

Chris Gammell: Yep. Yeah. Here we go. And his man's story too. Yeah.

Speaker ?: He's got a story.

Chris Gammell: He's listened to the story before. Yeah. That's great. I like that. I like that. You could just, you know, you could start to record that into the RetroPie and you could just have that, you know, play out of a, you know, a little PCM speaker, something like that.

Dave Jones: Oh man. Anyway.

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

Dave Jones: Yeah.

Chris Gammell: Well, we are in, we live in a time of abundance of, of processing for sure.

Dave Jones: You wouldn't know what else is in abundance. You don't like, like this segue.

Chris Gammell: Cabling in the mock. Was it? I knew it.

Dave Jones: He got it. He got it. He got it. Yeah. Oh, well done.

Chris Gammell: Wined that one up. And I, I spiked, I spiked to that baseball right there, folks. That's since this is an evening of sports, sports references. And yes, sports. I spiked to that baseball.

Dave Jones: Spiked to baseball. You spiked, they spike baseballs, do they?

Chris Gammell: They don't spike baseballs. I didn't think they did. Yeah.

Dave Jones: I thought I knew enough about American sports to know that wasn't a thing.

Chris Gammell: Yeah. All right. So it has an extra mile of wires. Tell me about this.

Dave Jones: This is a new Ford Mustang marquee. I haven't heard anything about it. I don't know what it looks like. I don't care. But yeah. I have.

Chris Gammell: It's, it's quite, it's, it's nice. It's actually, it's an SUV. It's like a crossover. You think Mustang, but they have a bunch of crossover model too. So that's interesting. Definitely hitting like that. Oh, I see.

Dave Jones: So it's a big ass sedan. It's like, it, it does look like a big ass Mustang with a big donk in the front, you know, even though it's got, right. It's a, um, is it fully? I mean, Mustangs typically were like a, uh, yeah, but it was like, Mustangs are typically

Chris Gammell: like muscle cars. Right.

Dave Jones: And this is like a muscle car. Right. It's got like a big donk on the front, even though there's just little hub, um, motors in the wheels, you know, maybe I'm looking at the wrong thing here.

Chris Gammell: The mock. Maybe. Maybe.

Dave Jones: Oh, no, I am looking at the, oh, what am I looking at?

Chris Gammell: It's a crossover, man. It's got a four door. It's the four door. Well, it's got, does it have a fourth door? Yeah. It's got a fourth door. Yeah. So it's like a crossover now instead of like, it still kind of has like that muscle car look, but it looks like it like a crossbreeded with a Tesla. No.

Dave Jones: No. Nope.

Chris Gammell: It's an SUV. Yeah.

Dave Jones: No, it's a sedan.

Chris Gammell: It's literally says SUV in the name.

Dave Jones: Oh.

Chris Gammell: The new 2023 Ford Mustang Mach-E SUV.

Dave Jones: Oh, okay. Right. All right.

Chris Gammell: Anyway. Okay. It's in the title of the webpage. Right.

Dave Jones: Well, the video on the page that we linked in is the muscle car. Like it just looks like a regular sedan muscle car. It's not an SUV. Anyway. Got it. Whatever. It's got an extra mile. What is it? Extra mile. How much extra cabling? How much extra cabling? Yeah. It's just nuts.

Chris Gammell: Yeah. I mean, harnesses are insane. I don't know if that would be an interesting episode if we had someone on here that was like, I don't know if they'd talk about it either.

Dave Jones: I've been meaning for years because I have the full teardown of the original Golf EV. I think I've got the full like 500 pair. The VW Golf? The VW Golf, I think. One of the original EVs is from like eight, 10 years ago or something. Anyway, I've got the full teardown manual for it, which itemizes every single part, every single connector, every single wiring. Oh, yeah. We talked about that. Yeah.

Chris Gammell: I thought it was a Beamer. I thought it was a BMW.

Dave Jones: It could be a BMW. Yes. Yeah. She might be right. Yes. Yeah. Yeah. It's a BMW. Yeah.

Chris Gammell: I remember we talked about that. Yeah. And that was like crazy because we did an episode about that sort of thing. Yeah.

Dave Jones: Yeah, we did. And I was going to do a video actually just going through and highlighting some of the amazing things in the teardown document. Like how insane, like the number of bomb items that go in a car is just mind blowing. It's just incredible. I think that even though it's a really ancient car, just the video would still be completely pertinent today. It's just nuts. Anyway. Yeah. So they've gone like, so how much is an extra mile of cabling? I don't know the exact.

Chris Gammell: In non-freedom units?

Dave Jones: Yeah. In non-freedom units. Yeah.

Chris Gammell: 5,200 feet. So divide that by three. So like 1,800 meters, 1.8 kilometers. Does that sound right? Let's just look it up.

Dave Jones: 1.6. A mile is 1.6 Ks.

Chris Gammell: Okay. There you go. Perfect. Okay. So 1.6 K. That's not a small amount. That's a lot, a lot. But I'm guessing they're also talking about bundled, right? So like, or if you like laid them end to end with the extra, you know. So if you have a bundle that's got, you know, four conductors in it, you're not going to say like, and the total bundle is like 10 meters long. You're not going to say I have 10 extra. You're going to, you know, if you don't need two of those conductors, you're not going to say you have 10 meters extra. You're going to say you have 20 meters extra. So I'm sure it's also stuff like that where it's like, they ran extra conductors for whatever reason and they didn't use it.

Dave Jones: Yep.

Chris Gammell: Or maybe they went really crazy and they unbundled.

Dave Jones: It would be very interesting to see like a modern car, like are they using like a bus? Because doesn't the average car, even average internal combustion engine car has like 200 processors in it? Like it's nuts. It's absolutely nuts. It's got a processor for everything. And are these all on a common bus? Like how does it, you know, I don't know.

Chris Gammell: Well, there's like the CAN bus and then there's like a Linn bus.

Dave Jones: And then they separate the drive buses are going to be different because if you're going to fire in the entertainment system, you don't want that to stop the car from functioning, right? You don't want the brakes to fail, for example. So, you know, they're totally independent.

Chris Gammell: Let me send you a, this is definitely an old photo. You can tell just from the photo itself, but here is a photo of an old harness that we can use as a reference point. I mean, just like the complexity of them is insane.

Dave Jones: Yep.

Chris Gammell: I'm not sure if that came through. There it is. And like, if you ever win the contract to like build a cable harness, it's worth so much.

Dave Jones: Yeah, that's a wiring harness board. That's a great photo. We can use that as a thumbnail for today's video. That's a great.

Chris Gammell: We don't own that. We don't own that image. Oh, come on. You know, you can maybe put in a cheat AI engine.

Dave Jones: I made this so different now. All right. Okay. Well, we'll use it.

Chris Gammell: I don't know if, I don't know if MidJourney knows what a cable harness is.

Dave Jones: Well, we'll try that after the show.

Chris Gammell: Oh yeah, we will. I'm sure. I'm sure.

Dave Jones: Yeah.

Chris Gammell: It's a chart this week.

Dave Jones: Anyway, for those who can't see it, it's like there's a, there's a production worker standing inside this giant six foot tall board, which is maybe 20 foot long or something. And it's got all these pegs coming out of it. And then it's got all the, uh, on the board. And then they use these loops to make the wiring harnesses. This is how they make them. They hand make them. Yeah. I assume they still hand make them these days. I mean, it's gotta be hard for a machine.

Chris Gammell: Oh, there's tons of automation, but it's not like, but I'm sure when you're testing it and building it out the first time. Oh yeah. That first build is crazy. Yeah. Yeah. I mean, it's just, it's a, it's a sub sub part of the industry for sure. I mean, like it's, uh, there's a lot of money in it and like, and like, if you ever have to, I remember, I remember hearing about like someone having to replace a cable harness, like a, like a animal got in there and started like gnawing wires and stuff like that. Okay.

Dave Jones: Yeah.

Chris Gammell: Yeah. And like, they just said, well, we'll just have to like replace the whole harness, but it was like thousands of dollars because of obviously labor too.

Dave Jones: And well, and that's not something you'd keep at a spare parts place. Like you wouldn't keep an entire wiring harness. You'd probably have to special order that. So yeah. Wow. Yeah.

Chris Gammell: Cheers. So if, if there is anyone out there, I'm not sure it would be like a super interesting show, but I'm sure there's tons of challenges in there. If anyone knows someone in that space, send them our way. We'd love to hear from them.

Dave Jones: Yep. Isn't there a dedicated Reddit? Isn't there a dedicated Reddit to what, like wiring looms and wiring harnesses?

Chris Gammell: I know there's that one that's like, uh, well, there's like that one for like a server

Dave Jones: cabinets where it's like, uh, it's like wiring porn kind of thing. Yeah. Something like that. Yeah.

Chris Gammell: Right. But that's probably different. I don't know. That's the thing. I don't think you ever like see, like, I feel like when you're a server cabinet, everything is visible and you need to kind of like troubleshoot it. But like, when you look at like how a wiring harness goes through a car, it's always like tucked into channels. You can't see them.

Dave Jones: Yeah.

Chris Gammell: Crazy. Yeah.

Dave Jones: You just can't see them.

Chris Gammell: That's also why it's so expensive to replace them. I feel like if it ever happened.

Dave Jones: You'd have to take half the car apart to.

Chris Gammell: Yeah, exactly. Like all of the finished fiddly bits that would be, you know, like crazy.

Dave Jones: And they've got so many things branching off them. So it's, it's not like, you know, it's not like there's a wiring harness with just two ends on it. No, there's probably, you know, 20 things branching off from it.

Chris Gammell: Yeah.

Dave Jones: You know? So yeah, yeah, it's a problem. Jeez. I really wouldn't like, I would hate to work on a modern car. It's just, I don't know how they do it.

Chris Gammell: Well, you wouldn't be doing it yourself. No, exactly. Yeah.

Dave Jones: Yeah.

Chris Gammell: But yeah, it's, it's a very complex machine for sure.

Dave Jones: There you go.

Chris Gammell: We will not be building our own. How's your, how's your electric vehicle going?

Dave Jones: It's great. I've done, I've had it for what, two and a bit years now. And I've done 40,000 kilometers in it, I think.

Chris Gammell: Oh, yeah.

Dave Jones: So it's actually quite a lot. Yeah. Yeah. It's great. Love it. Yep. Yep. Still almost entirely solar powered. Mostly. Oh, great. So yeah.

Chris Gammell: Great. Great. I actually, I posted something on the subreddit about, oh man, we have so many things here. Enphase has a new bidirectional EV charging.

Dave Jones: I saw that. And their marketing on that tweet is like, suits almost all EVs. And I'm going, no.

Chris Gammell: No, you have to have that built in, right? You have to have that thing where you can repower the house. Yeah.

Dave Jones: Yeah. You've got to have the vehicle to load, it's called.

Chris Gammell: Yeah.

Dave Jones: That has the vehicle to load capability. And no, not all EVs have that. Although I haven't.

Chris Gammell: Vehicle to house, V2H or V2 grid.

Dave Jones: Well, it's pretty vehicle to load, vehicle to grid, vehicle to house. Got it. It's all the same thing. Basically, you need to be able to export energy from the car. Power has to come out of the connector, not just in. And it's kind of a big deal.

Chris Gammell: What would the interlocks like on that too? Could you? I mean, I know they have plugs on them.

Dave Jones: In theory, it's possible to redesign. Oh God, I had somebody email me. I can't remember the exact details.

Chris Gammell: Always how it starts.

Dave Jones: And he emailed me and said, oh, look, I want to bypass the things on my car. Like I've got an EV and I want to like export. I want to power like loads from it and stuff. Somebody actually emailed me about this. And I'm going, dude, like don't, you know. And he goes, oh, I know how to fuse it properly. And like I'm going, dude, the amount of energy in that pack. Yeah. Like just don't. Oh, he goes, no, no, trust me. I know how to do it.

Chris Gammell: It's really too bad. I'm glad that Batman 2 is otherwise a terrible movie. Because like when at the end, the bad guy gets like completely like roasted by Selena Kyle.

Dave Jones: I don't remember that. If you know, you know. All right. Okay. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah. I'm going through the amount of energy in that. It's like I would so not want to touch it. Like, no.

Chris Gammell: Well, we are well on record that you and I are not the people to do that sort of thing. But yeah. Some people are just like, yeah, it's a big deal. It's just electricity. Yeah. And like some of them are like legitimate. I feel like, you know, who knows about that guy. But like some of them like actually just kind of have it in hand. Okay. Cool.

Dave Jones: Well, good for you.

Chris Gammell: That's not for me.

Dave Jones: Okay. Great. That's off. But I didn't want to borrow it. I didn't want to recommend anything. I just said. Yeah. Right. Exactly. Do it. You know.

Chris Gammell: Right.

Dave Jones: Anyway, speaking of solar powered cars, the. Yeah. Yeah. Aptira went bankrupt. I've done a video on the Aptira. Oh, what a loss. No. No one could have seen this coming. Who would have thought that a very ordinary EV with solar panels strapped on the roof that cost a quarter of a million dollars would have failed? Yeah. Yeah. Who would have predicted that? Oh, yeah.

Chris Gammell: It sounds like they were powered on dreams and, you know, the dream died. The dream died.

Dave Jones: And it's like, yeah. So I did. I did a couple of videos on that. And it's like, yeah. Sorry. Oh, sorry. No, not the Aptira. The Lightyear Zero. Sorry. I was completely wrong. Yeah, Lightyear Zero. The Lightyear Zero is this ordinary looking suburban, like ordinary looking four-door sedan, but it's covered in solar panels. And it costs like a quarter of a million bucks. And there was nothing special about it. Not like it was super performance. Yeah. It had a very slick drag coefficient and stuff, but whoop-dee-doo. Right? Yeah. It's like, you know, my EVs, you know, like you can buy one for like one-fifth of the price, one-eighth of the price, right, for a similar performance car. It's just crazy because they whack solar panels on the roof. And anyway, they've gone bankrupt. They're thinking about restructuring to only- Restructuring. Yeah.

Chris Gammell: Bring in the MBAs.

Dave Jones: Because they're going to, they said they're probably going to try and produce the, no, actually, no. Their first announcement was that they're going to stop producing the Lightyear Zero, and they're going to produce the cheaper, which was the quarter of a million dollar one. And they're going to actually focus on making the, you know, $50,000, like the much cheaper Lightyear One.

Chris Gammell: All right. Because that's easier. Yeah. That's the easy part. That's much easier. Sure. That's much easier.

Dave Jones: So I actually tweeted when they made that announcement. I reckon they'll go bankrupt shortly. And they didn't even last a week before they announced bankruptcy. So I totally called that one. Yeah. Because it's so easy just to switch from a quarter of a million dollar low volume production car to a high volume production.

Chris Gammell: Yeah. The high production is really easy too. Bootstrapping. Yeah. Yeah. Good lord. Yeah.

Dave Jones: Bloody hell. That's like, come on. How does that like, well, I guess there's rich people who just want to say, I've got a solar powered car. Right. And they're willing to pay a quarter of a million bucks for it. I don't know.

Chris Gammell: Well, if they are, you know, those people would also love our solar powered sponsorship of The Amp Hour. So please get in touch if you'd like to sponsor the solar powered Ampower. How would that work? Less than a quarter million, but more than 50,000. Yep. I don't know, Dave. I'm just trying to get free money.

Dave Jones: Okay. Sounds good.

Chris Gammell: Yeah.

Dave Jones: Let's milk that, let's milk that solar cash cow.

Chris Gammell: That's right. Yes.

Dave Jones: Milk that eco-friendly cash cow.

Chris Gammell: Oh my God.

Dave Jones: Yeah. There has to be a government subsidy in it somewhere.

Chris Gammell: Come on. You're in America.

Dave Jones: There's got to be. There's got to be a subsidy somewhere in America.

Chris Gammell: I'm sure there's subsidy, but no, that sounds like a European thing to me personally. Right.

Dave Jones: Okay.

Chris Gammell: There's no, you know, there's this corporate greed here. There's not so much on the, you know, the subsidy side of things. I do think, you know, like there's all this greenwashing stuff. I think it may be mentioned on the show, but I was talking to a buddy who like, I was telling him about this IOT heart, the trash can demo that I built. Right. For my work.

Dave Jones: Oh yeah. Right.

Chris Gammell: And he's like, oh yeah, actually I know someone who's doing that. And not only, and they're doing like the operation side of it. And you know, he was like, oh yeah, maybe they would like this sort of thing that I had built. And, um, but not only are they like selling the service to a municipality, right. They're like, actually like being able to like monitor trash cans. So the trucks don't have to go out as often. And then they're then, then they're able to sell the fact that there are fewer trash trucks on the road for like these, these like super squishy green credits. And I just like, I, I don't know where this, this stuff's going. There's a, there's a, um, last week tonight episode, the John Oliver, um, show where they talk about these credits and they just seem so scammy to me.

Dave Jones: And I just, I'm sorry if people. Is it Wendover productions who did a video on carbon credit? Did he do one too? It was Wendover. It was one of the, I think it was Wendover.

Chris Gammell: Wendover probably did something like that. But I always think of the John Oliver thing. I'll link if I'll look for the Wendover one too, but yeah. I just, it just seems like these like, like fun, these fun notes, you know, like it just doesn't seem like there's nothing behind them. There's no like enforcement. There's no like, and it, people are just selling like literally thin air, you know? Yeah. Right. Yeah. Yeah. Here we go.

Dave Jones: Yes. It is. The Wendover production video. I'll send you a link. Okay. I'll like that.

Chris Gammell: Yeah.

Dave Jones: Yeah.

Chris Gammell: Yeah.

Dave Jones: And, um, yeah, I don't, yeah.

Chris Gammell: I just feel like there's going to be some kind of reckoning in the next couple of years about it because. Yeah. It's just, I, I got bad feels from it. You know what I mean?

Dave Jones: Well, another thing that's a scam is the, um, ESG thing, the environment. Well, I don't know what it stands for. The environmental, uh, yeah. Companies are supposed to have this environmental score and it's this, you know, it's this, yeah. And if you don't have a low enough or a high enough score or whatever, then you don't get money from BlackRock or you don't get money from, you know, somebody, you know, JP Morgan or somebody like that. Right. And yeah, it's, it's complete bullshit. It's complete. Yeah. Yep.

Chris Gammell: Uh, you know, it's interesting. Yeah. I did. I think I heard a podcast about, I didn't. Recognize the acronym, but there was a podcast about like the, that, like whether that is kind of scammy or whether it's like, uh, whether it's driving behavior.

Dave Jones: ESG stands for in environmental, social and governance score. So you're, you're, so your company is rated on your environmental score, your social activism score and your corporate governance score or something. And it's just, yeah, it's complete bullshit and it's failing epically.

Chris Gammell: Um, it's going down the toilet.

Dave Jones: It's basically virtue. Yeah. They rate you on your virtue signaling ability and, and how many carbon credits you buy and all that sort of wankery.

Chris Gammell: So, yeah, I mean, like that kind of thing would drive like, and that's the thing, like, I feel like all these companies have like these, uh, these mandates to do something. Right. And they're like, they're trying to reach out and they're like, oh, look, this guy won't cut down his forest in the middle of Pennsylvania. Yeah. Right. Okay. But, and, and like, they kind of feel like they have to and like, okay, you know, they're going to do it. Yeah. Yeah.

Dave Jones: Something got sucked into all these programs and things that people offer, you know? Yeah.

Chris Gammell: Yeah. Yeah. Yeah.

Dave Jones: Yeah.

Chris Gammell: Yeah. Yeah. Um, what was the last thing? Oh, uh, speaking, speaking of, uh, carbon credits and, uh, lack thereof, I will be flying to Germany, um, uh, which is, uh, you know, not a small amount of carbon, uh, but I will be at a better world and past guests of the show saber from PCB arts. They will be doing a meetup there on the Wednesday night of that event. And so if people want to go and meet saber and see the vapor phase one and, uh, hang out with other nerds like yours truly, uh, cool sign up. So, and if other people are at embedded world, love to hear from you as well. Um, yeah, I'm excited. I'm on a sprint until then.

Dave Jones: Are you doing anything over there or are you just going for the duration of the show?

Chris Gammell: Just the show. Yeah.

Dave Jones: Bugger. So this is a corporate junket.

Chris Gammell: Uh, what are the other options?

Dave Jones: Uh, you, you're doing it off your own bat and you're paying for it yourself. Oh yeah.

Chris Gammell: No, this is corporate. Yeah. Right.

Dave Jones: Okay.

Chris Gammell: Yeah. I'm going to like showcase. I'll be at the Zephyr booth showcasing like I was last year. Um, yeah, it's a good time. I just really good show. So I, I highly recommend it. Uh, if people haven't been, I'm not sure. Like, so like if I was advising consultant, Chris, whether to go, I don't know. But if I was insulting, like if I was consult, uh, I was recommending to like Chris at past job and I could convince Chris at past job to, to go. Yeah. I think, I think it's one of the best trade shows to go to.

Dave Jones: If you're going to go to any trade show. I would totally love to go, but you know, it's on the other side of the planet. Oh, so far. Kind of expensive. Yep. Yep. 30 hours of travel. Just to. Yeah. Yeah. Yeah. Anyway, there will be a prize for the, uh, said, uh, Chris meme photo. So if you can get a photo of said Chris sitting on the booth, just looking like you got to, you got to, right.

Chris Gammell: You, you know, like you, you have to actually list the prize. I feel like you never really do that.

Dave Jones: I've got a list of prize. Let's got to be like a free multimeter or something for said Chris. Yeah.

Chris Gammell: Yeah.

Dave Jones: All right.

Chris Gammell: That's the thing. I'm full of joy. I couldn't, you couldn't catch a bad picture of me if you tried.

Dave Jones: Oh, come on. Happened to poor Keanu, you know, happened to sad Keanu.

Chris Gammell: Sad Keanu. Eating a taco. Yeah. Yeah. You might catch me actually. The vendor solutions are, um, you know, sorry, the food solutions are not, uh, great. Oh, man.

Dave Jones: The food always sucks at trade shows. Yeah. It's, it's the worst. Like I, I, like sometimes I just like starve myself. I just go, no. Yeah.

Chris Gammell: I just got to bring a, bring a box full of protein bars.

Dave Jones: Yeah. Protein bars or something. Yeah. Yeah. Totally. They're always awful.

Chris Gammell: Yeah. I mean, pretzels at least in Germany. So that's nice. Um, I think that was the only other thing on the, uh, the list that I wanted. Obviously lots of great links from an unmanaged 16. Once again, thank you for that. Yes. Thank you very much.

Dave Jones: But often we don't get around to going through the links cause we just waffle on a lot of random shit.

Chris Gammell: I mean, there's a lot of great stuff in here.

Dave Jones: Yeah.

Chris Gammell: I mean, we got to get our stuff in too.

Dave Jones: For those who don't know, we don't plan the show that there are no notes. It's just, we just press record and go, welcome to the amp hour. And then, I don't know.

Chris Gammell: Sometimes not even that. Yeah.

Dave Jones: Something comes out. Yeah.

Chris Gammell: Yeah. Yeah.

Dave Jones: Yep. Hang on. Has Tesla dropped the AM radio? Have they?

Chris Gammell: This is, this is what grabbed your eye Dave. Yes. I am.

Dave Jones: Well, you know, I, I bet you can't even remember the introduction of AM stereo.

Chris Gammell: AM stereo.

Dave Jones: No, I am stereo was a thing, dude. It was a thing.

Chris Gammell: When did that happen?

Dave Jones: Oh, eighties sometime.

Chris Gammell: Early? I don't know.

Dave Jones: I have to look it up. I actually had an AM stereo receiver.

Chris Gammell: How about this? I would not have been in charge of the radio dial.

Dave Jones: Hang on. I am live Wikipedia in this. I could have, mid, mid to late eighties. There you go. There you go. I was on.

Chris Gammell: Same, same response from me. Yep. AM stereo.

Dave Jones: It was all. It was all the rage for about a month. Okay. Yeah.

Chris Gammell: Interesting.

Dave Jones: Yep. Anyway.

Chris Gammell: This is not what they're getting rid of in.

Dave Jones: Oh, aren't they?

Chris Gammell: In the Tesla, but.

Dave Jones: All right. Okay.

Chris Gammell: It's crazy too, because you think of like the chipsets that have, that are tuning in FM. They have AM, like, I guess it's probably just the, the, the space for the antenna and stuff too. They like have it.

Dave Jones: Well, they're different architectures too. Like, not only do you need the, well, you've got to have the rod antenna, right? You've, you've got to have the, you know, the ferrite rod for the AM and everything. Like, I don't think you can get away with it. Can anyone do a AM receiver without the ferrite rod? It's, I don't think so. It's such low frequency that.

Chris Gammell: Hmm. Yeah. Yeah. I'm thinking, uh, yeah, here we go. EDN fully automotive AM FM radio receiver. I see reduces count by 40%. So like, there is like, there are chips that do both, but yeah, I guess I think it would be the other stuff that's in it. Like, you're right. I think the antenna is probably much bigger. So that's probably not worth it to Tesla, but I'm pretty sure the, the chipset is like super commoditized at this point. Like you can buy, you can buy off the shelf. I'm sure. Like receivers that are, they just pump that stuff out. So.

Dave Jones: Yeah. I haven't looked at God, an AM FM chipset in 25 years or something. That's how long.

Chris Gammell: Today is that day.

Dave Jones: All right. Chip of the week.

Chris Gammell: Bingo.

Dave Jones: AM FM transfer chip.

Chris Gammell: This article I sent you was from 2008. So I don't know if chip of the week is really it, but you know. Yeah. Yeah.

Dave Jones: Oh boy. Yeah.

Chris Gammell: All right. We should end the show though. End of show. Now. Ish.

Dave Jones: Ish.

Chris Gammell: All right.

Dave Jones: Catch you next time.

Chris Gammell: See ya. We'll be right back.