#396 – The Synergy Bus

Chris Gammell: This is The Amp Hour Podcast. Released June 10th, 2018. Episode 396. The Synergy Bus.

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

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

Dave Jones: What's up, nerd? Hey, buddy. How you doing? Doing all right.

Chris Gammell: Okay. Apart from the usual crap. Yeah? What's the usual crap?

Dave Jones: Ah, nothing.

Chris Gammell: I'm just wondering if you're going to say it or not.

Dave Jones: No, no, no. I tweeted it, you know.

Chris Gammell: Dave's getting older, folks.

Dave Jones: I'm getting older. Dave's getting nerdier, folks. Should we get it out of the way now?

Chris Gammell: Yes, we should.

Dave Jones: Yeah, all right. Yep. Apparently, I need glasses.

Chris Gammell: Oh. I mean, like, I guarantee, like, 80% of our audience. Come on, we're all nerds here.

Dave Jones: Yeah, I know. It's like, every single person I know has glasses. And they're all, like, most of them are younger than me. You know? It's like, you know. I do.

Chris Gammell: I wear contacts.

Dave Jones: I know. So, yeah, like, yeah, you've probably heard me, you know, complaining about it on the blog. Oh, I can't see stuff up close, you know, anymore. Like, I'm talking, like, within, you know, 30 centimeters, you know, 12 inches for you, Yanks. Yeah. Kind of all that sort of closer. Yeah. And it becomes a bit harder to focus. Yeah, yeah, yeah.

Chris Gammell: I always like the, there is someone. Oh, it's in White Christmas. It's a Christmas movie. Right. One of the guys is talking about, like, I'm starting to play trombone. And it's talking about, like, when he has to, like, move the paper in and out from his face to actually focus in. He's talking about playing trombone with Bing Crosby. Right. Yeah. It's a part of life, Dave. Yeah. What's interesting there, I think, is what we were talking about before the show is, I think, you know, we were, I was obviously expounding upon all of the tech that could help people as they're getting older, too. I think that's a huge, a huge opportunity for people listening. Like, if you're thinking about starting a business, you're thinking about, like, just baby boomers. I mean, Dave, you're younger than baby boomers, but, like, people, as they're getting older, I think assistive technology type stuff is going to be a huge thing. And I don't think much would help you other than, you know, well, I guess classes are assistive technology. But I think that, I think there's just a huge, a huge industry there. And just to keep an eye out for it, you know.

Dave Jones: Definitely. No, it's no doubt.

Chris Gammell: I've, we've talked about, obviously, self-driving cars on the show and your lack of, you know, confidence in them and me coming around to that point of view lately. But I think that, like, that is a huge one. You know, if that does, you know, in 10 years, hopefully that'll help my folks. Like, they're, they're kind of, that'll probably be about the point when they won't be able to drive much anymore. So, you know. And honestly, I won't want them to drive.

Dave Jones: But then you've got to fight the crusty old, man, it's newfangled technology, you know. Yeah, I talk to you every week, man. I mean, yeah.

Chris Gammell: Yeah.

Dave Jones: Which you wanted to talk about. Speaking of newfangled technology. Tell us where you're going and what crusty old technology you're going to be using. I'm very excited about this.

Chris Gammell: Yeah. This is great. So, I think I've mentioned, I will be at two, two events this summer are my main two travel things. First one is Tor Camp, which is coming up in two weekends. And then I'll be at DEF CON in August. But at Tor Camp, I found out today, because I just didn't read about it, there's this, there's this thing called Shadytel. And, and it's only going to be, it's not doing GSM service this year. So, in the past, they've had like a, like a, like a, a cow or like a, a portable, you know, antenna service where you could put a SIM in. Yeah. That's not what they're called. Sorry.

Dave Jones: Oh, it's an acronym, is it? Or something. Yeah, it's an acronym for like a portable tower. Anyway, one of those portable cellular leaks. Yeah. Exactly. Yeah. Right.

Chris Gammell: They call them stingrays now too. I mean, the, the government ones that are like spying on people are called stingrays, I think. Oh, okay. Go ahead. But yeah, so the stingrays or cows or whatever, but basically those could actually do off-brand GSM service with like special SIM cards. They won't be doing that this year, but what they will be doing, and apparently have done in years past, is they, they wire, they wire cabling to every tent so that you can bring an analog telephone and plug in an R, was it RJ?

Chris Gammell: 22, 24? What's the, the jack type on old telephones?

Dave Jones: RJ 11. Well, no, no, no. That's no, RJ 11 is that modern rubbish. No, no, you're talking, God, I can't remember the, the, the three. I don't know whether it's the same in the US, but I can't remember, like 601.

Chris Gammell: Uh-huh.

Dave Jones: I have a vague recollection of. Anyway, yes, the old fashioned telephone plug. Okay.

Chris Gammell: Yeah, the RJ 11 is the one I think of like with the gray cabling. Yeah, yeah, of course. And it's got the, the, usually the clear plastic end. Yeah. I've also, the, the acronym today. That's great. And I'm going to have a fobs who's been on the show before and I'll link to that episode. So Fabienne was talking about, she's going to be up there. Um, uh, she was tweeting about it today and this is kind of how I found out about it too. So all of the things that are happening, but the one that caught my eye was the Potts phone. And I was like, what the hell is Potts? Do you know Potts?

Dave Jones: Potts.

Chris Gammell: No? Yes? No? Point to yes. Plain old telephone service.

Dave Jones: Plain old telephone service.

Chris Gammell: Yeah. I didn't, I didn't know that one. So yeah. So basically they're going to wire, send right, wire each tent.

Dave Jones: This is great. Each tent. And it's not a party line, which would be the easy way, but it actually goes into a little local exchange and you can dial a tent. I think so.

Chris Gammell: Yeah. Yeah.

Dave Jones: That is awesome. So Uncle Sam's not going to be listening. It's great.

Chris Gammell: Yeah.

Dave Jones: As, as I was saying before the show, if you're going to be paranoid, go all the way.

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

Dave Jones: That's great. Yeah. That's great. So there'll be no cellular coverage there.

Chris Gammell: Uh, I think, uh, I looked at the map today. So actually because my, my job, uh, I know about service, this is actually a pretty, pretty good resource that people might want other, other days of the week. It's called open signal. Uh, and it's a nice mapping service. I think people basically war dial, like you can get an app and you can opt in to, to help give coverage maps. Uh, and it just checks in regularly. So I was looking at it today for Orca Island, which is where the, um, where the, the conference is going to be. And there's stuff like on the shore, but definitely inland from there. It's not going to have much. I don't know where the tower is. Right. Um, so yeah, um, I'm, I'm really looking forward to it and it's going to just be a nerdy weekend. So I'm going to, you know, be bringing projects. I'm going to actually be giving a Kaiket workshop. Um, that's kind of interesting. Cause I'm gonna have to bring, they said there's internet as well. I'm not sure how they do the internet stuff, but, uh, uh, I'm figuring that at a hacker

Dave Jones: camp, it's going to be like pinging through five anonymous routers. Oh yeah. You know, shit like that. You know, somebody's going to take that by the balls. Turn your VPN on either way, right? Yeah. Yeah. Someone's going to take that by the balls.

Chris Gammell: This seems like a very friendly camp.

Dave Jones: No one's getting in here, you know? Yeah. Yeah.

Chris Gammell: Right. But I don't, it's really stupid too. Like, like I'm like really looking forward to like having like some quiet time to like do coding and to sit down with people and do some ham radio stuff. Like, cause you know, like I don't really get a chance to sit down with people when they're using their radios, but just like watching people, um, use ham radio stuff. Um, looking forward to that. Uh, maybe do some hack RF stuff and do some SDR. Cause I think that Hulk, I think some of the great Scott gadgets folks will be up there. Um, and you know, just learning from people and camping and making marshmallows and s'mores and s'mores and nerds, you know?

Dave Jones: S'mores, sorry. Oh, you don't know s'mores.

Chris Gammell: Oh, you're killing me, Smalls. What? Uh, you ever, you ever seen, uh, uh, the Sandlot? No?

Dave Jones: No.

Chris Gammell: Oh, okay. No. You have homework now, Dave. Uh, but anyways. You're really losing me today. I know.

Dave Jones: You were throwing shit at me before the show and I'm just going, nope, nope, nope, nope. Yeah, that's okay. Nope. No idea.

Chris Gammell: Uh, so s'mores are when you roast marshmallows and then you put them onto a graham cracker with a chocolate, piece of chocolate.

Dave Jones: I don't even know what a graham cracker is.

Chris Gammell: Wow. Man. We live so... Dave, the chasm is so deep between us. Uh, yeah. First you take the mallow, then you roast the mallow. What? Anyways, I will, I will, I will post a link and send you a link afterwards. Right. Okay.

Dave Jones: Thank you.

Chris Gammell: Yeah. Man. You thought the glasses made you sound old. Right. Oh, man. Yeah. Yeah. Um, so yeah, I'm, I'm looking forward to it. I mean, like...

Dave Jones: Do you think, do you think that the government pay or any three-letter division of the government pay any attention to these camps at all? Do they try and, like, infiltrate them? Do they try and hack them to see what's going on there? Do they... Oh, uh... Or do you think they just don't care? Come on. Conspiracy theory type. I don't think it's big enough. But, like, why...

Chris Gammell: It's not big enough? Why would they? I mean, like, maybe... I would be just more likely... Why would?

Dave Jones: Because all the talks are published online afterwards, aren't they? Or something?

Chris Gammell: Well, yeah, that. But, I mean, like... But at that point, they could just attend... Like, so... Like, a big component of people that go to, like, DEF CON and other security conferences are working for government agencies. Oh, okay. Right. Yeah. Yeah, yeah. Got it. So, a lot of them, like... A lot of people who go to those things won't talk about their work simply because they're not allowed to. Okay. Yep. Yeah, I was surprised by that last year. I think they said... I think... So, DEF CON is in August. But there's B-sides, and then there's also Black Hat. Black Hat's more... It's owned by UBM now. But I... And it's very, like, corporate level.

Dave Jones: Oh, that sounds like the death...

Chris Gammell: No, it's been going for a while. It was bought, like, 15... 10, 15 years ago, I think. Yeah. Okay. So, it's been around for... It was not started by them, but it was bought, and it has continued on. But that's actually, like, a, you know, pay a couple thousand dollars to go to it kind of thing. Got it. Yeah. So, there are those kind of things and those kind of conferences, and I think that they continue to exist because there's, you know, the quasi-academic nature of them and the... Right. Just the cost of access kind of... Yeah.

Dave Jones: Speaking of buyouts, you heard that Microsoft bought your favorite GitHub.

Chris Gammell: Yes. Yeah. Are you sad? I don't really care. You don't care, right?

Dave Jones: As long as it doesn't change, right?

Chris Gammell: I mean, like, honestly, I just thought I'd think about that. There was a tweet about it. It was like, all these people are upset that Microsoft buys, you know, the centralized service is buying GitHub, but at the same time, it's the same people that took a completely distributed architecture and paid a singular service to host that for them.

Dave Jones: Yep.

Chris Gammell: Right? Like, when GitHub goes down, everyone freaks the hell out. Right. You know? And that's not what Git's meant to be. It's meant to be like, oh, you have a copy, I have a copy, there's other copies.

Dave Jones: Oh, really? It's supposed to be a distributed service. It's distributed, yes. Okay. Right. And that's why, like, so... And that doesn't work?

Chris Gammell: Well, it does work, right? So if you have a copy and I have a copy and we both pulled from GitHub, it's a centralized copy. But GitHub, in that case, is the official copy, right?

Dave Jones: Oh, okay. Got it. Got it. Yep. And then once it comes back online, everyone sorts themselves out, you know?

Chris Gammell: Yeah, exactly. Right. Exactly. And there's, I mean, there's other services built on top of it, too. Like, I think certain things...

Dave Jones: Well, that's the same as Dropbox, you know? Right. If Dropbox goes down, I don't care because I work for my local copy, you know?

Chris Gammell: Right. And I think that that's, you know, I think that's the services that I gravitate towards these days in the first place of, like, I don't want... You know, well, we've talked about cloud only on here quite a bit, right? Mm-hmm. From a PCB perspective or EDA perspective more specifically. And I think, and I think you've said similar, but, like, if there is a local cache copy, then that's usually fine, right? I mean, like, you know, maybe we do or don't want to pay for the service, but I think that people usually recognize that paying for that centralized service and the updates and the whatever, that's what they're really charging for. Or if you have to also, though, have access in order to even get to the design files, that's problematic, right?

Dave Jones: Hmm.

Chris Gammell: No? Disagree?

Dave Jones: I intend to... No, it's fine. Okay. Yeah.

Chris Gammell: So, I just think that... So, like, Evernote's another good example. That's what I keep notes in. Oh, yeah. Yeah. I just like having, you know, if the service is down, whatever, or if I'm on a plane more specifically... Oh, definitely.

Dave Jones: Definitely. Yeah. No, you want to work from your local copy and seek later. Right. Right. Exactly.

Chris Gammell: Yeah. So...

Dave Jones: That's a no-brainer. Who wouldn't? I mean, you know... I don't... Yeah. You know... I always lamented that... I don't understand why you wouldn't use that.

Chris Gammell: Trello was the one that I always wished that they would do that. So, Trello is kind of like a... It's kind of like replicating Post-it notes online, but...

Dave Jones: Well, why do you use it, then, if it doesn't? Find one that does. I stopped using it. No, I stopped using it.

Speaker ?: Oh, okay.

Dave Jones: Right. Yeah. For that reason, or for... Yep. Exactly. Combinator. Oh, okay.

Chris Gammell: So, like, the Android app allowed you to work offline, but when I was on an airplane, and I wanted to keep notes on things, I just... I could... It just wouldn't work, you know? Right. And I think some of it is just, like, how the web technologies are, I think. I don't actually know. But they didn't build it like that, so it didn't work for me.

Dave Jones: Hmm. Interesting.

Chris Gammell: Yep. As for Microsoft buying GitHub, I don't... Eh. You know. GitHub is getting pretty big, too. You know? It's not like it was some tiny, tiny thing.

Dave Jones: No, no. Yep. That was enormous.

Chris Gammell: Yeah. Yeah, it was bought for seven and a half billion, I think.

Dave Jones: Yeah, it runs half the world's coding, doesn't it? I mean...

Chris Gammell: I don't know about that, but...

Speaker ?: Yeah.

Dave Jones: Oh, it has to be enormous.

Chris Gammell: I do... Here's what I do know. I know that when people are like, yeah, I have a Git instance, and I'm like, oh, yeah, GitHub. They're like, no, no, no, I host it myself, or I have other infrastructure. Right. Oh, you've got a lot of server. Right? You've got your own, yeah, yeah. Yeah. And it's like, totally, that's how it's supposed to happen. That's totally possible. Right, that's how it's. Yeah, yeah. Yeah, I'm used to this GitHub model now, and...

Dave Jones: So, large corporations would have that. They'd be running their own local instance, you know? Yeah.

Chris Gammell: So, like, Atlassian is like a big Australian software company, right? Yep. They'll... What is it? Bitbucket, I think? Is that what is it? What is their... No, not that Bitbucket. They have another, you know, repository-type service that you can buy and host somewhere. Okay.

Dave Jones: Yeah. Right.

Chris Gammell: At the end of the day, the thing I like about any of these services, and, you know, to talk about Dropbox, and, like, you're talking about, or Git, or whatever, at the end of the day, I just like having tracked points. Right. You know, auto-saves one thing, but actually having, like, either, like, the services backing itself up, and I can go and pull one of those things based on date, or my preference is to actually be like, oh, this is the change I just implemented, and actually writing that down and saying, I just, you know, I just changed, I just added a bunch of new resistors to this schematic. You know, capture that. Yeah. That helps a lot.

Dave Jones: Groovy.

Chris Gammell: Yeah.

Dave Jones: All right.

Chris Gammell: All right. Speaking of hardware, I've been... Speaking of hardware, what have you been doing? Yeah, speaking of things that I've been adding to schematic pages, I've been working on a new layout, but I've been having some issues with I2C. This is another embarrassing...

Dave Jones: Dude, it's two wires, two pull-up resistors. How much trouble could you have? Tell us.

Chris Gammell: Famous last word, yeah. Um, I had never really, even though I'd worked with a bunch of sensors, uh, in the past, they were either serial, or one wire, or spy, I had never really worked with I2C before. Yeah, well that's... Or they just worked, you know?

Dave Jones: Right, right. Yeah.

Chris Gammell: Got it. And, uh, and so I've been, I've been in the throes of the I2C.

Dave Jones: You've been troubleshooting. Did you have to pull out the Bible?

Chris Gammell: I actually don't even know what the Bible is.

Dave Jones: The standard, like, the I2C standard. Did you have to get that desperate?

Chris Gammell: Oh, oh, God, no, no.

Dave Jones: To make sure it's meeting. All right. Okay. No.

Chris Gammell: Um, mostly, you know, like, my troubleshooting process was mostly around, like, first off, just checking boneheaded things, like...

Dave Jones: Like, make sure there's a signal there.

Chris Gammell: Well, yeah, or even before that. Make sure it's transitioning, right? Yeah, I was just looking at, like, is the code outputting anything to the lines, right? Oh, right. Yeah, I guess that's signal there.

Dave Jones: Yeah, you scope them, yeah.

Chris Gammell: Right, right. So I was, I was playing between different platforms, too. So I've been on the ESP32, and I was having a bunch of troubles with the, I was using MicroPython on it, and I was having trouble with that, and then, so then I ended up using Arduino and pushing code through that to the ESP32, and I probably should have taken...

Dave Jones: Oh, as a test platform, you would do it, like, as a debug platform?

Chris Gammell: No, well, so I had, so that was another problem. I had, I have an ESP32 on board. Yeah. But that was giving me other problems, and so what I did is I took a known ESP32, like, the Huzzah32, which is a Adafruit feather form factor, and just compiled for that. Right. So, put in Arduino IDE, had, like, a test program that I just, you know, basically was using the wire library, and then...

Dave Jones: So you couldn't get iSquadC working on an Arduino? Arduino.

Chris Gammell: Well, I could eventually, but that was, that was actually after, like, I was frustrated and trying to do it, like, all in one on this one platform that I had been building, and then I was like, no, that's stupid. I need to, like, make sure that my signal generator is working properly. Okay.

Dave Jones: So you were using a hardware iSquadC module?

Chris Gammell: No. What do you mean by that?

Dave Jones: Like, as in, like, the iSquadC hardware is built into, you weren't doing software bit banging?

Chris Gammell: I was not directly doing bit banging. Right. However, on an ESP32, I found out that it actually is a software implementation of iSquadC.

Dave Jones: Oh, it is. Okay. Yeah. Got it. Yeah.

Chris Gammell: So, and I was surprised by that, too. Mm-hmm. Especially because I was looking at, well, so, like, you look at the huzzah 32, right? And that's the feather form factor, which I like. And on the board, it says pin 32, or pin 23 is SDA, pin 22 is SCL, so that's data and clock.

Dave Jones: Well, if they're tying them to lines, then it's most likely using a hardware module. Right. That's why they tie them to lines.

Chris Gammell: Right. And what I found out, though, is actually that's because the feather form factor on, so, like, regardless of which, so this is Adafruit's form factor, right? Mm-hmm. Regardless of which processor they put on board, though, they will tie the iSquadC lines to those two pins.

Chris Gammell: To those pins. Yeah, yeah. Right, right, right. Right. Gotcha. And so what I found out, though, is that if you go and trace that back to the ESP32, you can actually instantiate iSquadC on other lines, meaning that it's software-based. Yep. Well, not necessarily. Not necessarily. No, you're right. You're right.

Dave Jones: It could have internal routing. You're right. You're totally right about that. Yeah.

Chris Gammell: I misspoke there. But I did find out it's software-based, so. Right. Okay. Right. Right. All right. Because I had been worried about that, too, because I was like, I've actually got three different iSquadC devices on board, which also complicated my design. I was trying to do, like, way too much at once.

Dave Jones: And they're all on the same bus?

Chris Gammell: They're all on the same bus.

Dave Jones: Got it.

Chris Gammell: And I hang the bus off to have other things plug in.

Dave Jones: Right.

Chris Gammell: This is for my DEFGON badge. So, yeah. Lots of people. If people are out there listening, there's the shitty add-on, it's called, which is, like, an ad hoc standard that Mr. Brian Benchoff came up with. Um, basically, I'm not sure why he chose iSquadC, but he put iSquadC as the ad hoc standard, basically. So, it's power, ground, and SDA, SCL, four pins, and this basically four-pin header that people are now making boards for that you can then go and plug into random badges. I think there's going to be people messing up, uh, iSquadC lines, like, if you have other things on them. But I have actually now, next rev will have separate lines for those things.

Dave Jones: It's not hard to upset an iSquadC line.

Chris Gammell: It is. Yeah, exactly, right?

Dave Jones: So... You do know the golden rule of troubleshooting iSquadC lines?

Chris Gammell: Uh, thou shalt check pull-up resistors.

Dave Jones: Thou shalt lower the value of their pull-up resistors. Aha! Good rule. When in doubt, lower.

Chris Gammell: When in doubt, throw it out.

Dave Jones: The standard, i.e. the Phillips standard, calls for 2k2 as a nominal reference pull-up value. But when you've got multiple, when you've got longer lines, higher capacitance, and stuff like that, you want to be, you know, really. Because if people don't know, it's not a totem pole driver. It's an open collector driver system. So the pull-up is what pulls the signal back high. And if you've got a lot of capacitance on your line, it takes time to charge up and you don't get a nice sharp edge and all hell breaks loose. Right. So, yeah. You want a lower value resistor.

Chris Gammell: So a lower pull-up resistor basically dumps more current directly into that node right at the receiving device. And it pulls up faster. Yeah. Right. Yeah. Yeah. Which is crazy. Because I always think about it like kind of like, like an I squared C line, to me, I always kind of visualize it like if you're on a bus and you're pulling like the stop line. Have you ever seen that? Pulling the stop line? There's like a line, like the bell to stop the bus. Usually there's like a wire that goes there. And if you think about it, like the actual actuator is usually at the front of the bus and then you're actually literally tugging down on the stop, the stop cable, I guess it would be called.

Dave Jones: That's not a bad analogy.

Chris Gammell: Thank you. Yeah. But the idea is that like the propagation to get from my arm pulling down on the line to actually activating the other end, there is some delay there. And obviously it needs to have some kind of springy force to go back up.

Dave Jones: That's right. Yep. Yeah.

Chris Gammell: And so the springy, a spring to make it springier. Yeah.

Dave Jones: That's why people slam down on that thing hard and fast.

Chris Gammell: That's right. Yeah. Right. Got to make sure you stop. Yeah. Yeah.

Dave Jones: They've got a lower resistance. There's just less propagation delay when you pull it harder and faster.

Chris Gammell: Right. Right.

Dave Jones: Yeah.

Chris Gammell: Yes. Nothing like an analogy. Yeah.

Dave Jones: I love I2C. I've written my own I2C library. Yeah. You said that. Because as we were mentioning before the show. Yeah. I was actually, I had an idea for a product. Like this is way, way back. We're talking, you know, could be 20 years ago now. Right.

Chris Gammell: So this was probably like a display, like a two line display with like a PIC controller or something.

Dave Jones: It was like a two line display with a PIC controller back before microcontrollers really had I2C modules, you know. So.

Chris Gammell: No, that's a good reference too though. Yeah. Yeah. I think that's, that's an interesting data point too. Like you said, the Philips standard, but like they made it the standard and they actually branded the standard as well. And now we just talk about it like it's a, it's a Kleenex, right? Type of thing where the brand name.

Dave Jones: You can't, to this day, I believe you still can't use the name I2C. It is trademarked by Philips. If you want to use it, you have to pay them a royalty. Interesting. You know, so that's why they call it 2Wire. Yeah. I think we've talked about that before actually. So. Yeah. And, and if you wanted to use the, I call it I2C and use it, it might've been free. There might not have been no royalty, but you had to get it officially approved by Philips to use the logo. If you wanted to use the Philips I2C logo, then it had to be, you had to submit your chip, your whatever it was to them. Yeah. So they evaluate and go, yep, this meets the standard. Congratulations. Here's the big stamp. You can use I2C. Oh, here we go. So I'm looking at the wiki page.

Chris Gammell: Not, no secret source here. Since October 10th, 2006, no licensing fees are required to implement it. However, they are required to obtain a slave address allocated by NXP.

Dave Jones: Ah, there. That's how they get you. I was about to talk about that. Ah, the old school slave addresses. Right. Because there were, once this system caught on, right, they used to use an 8-bit address so they could have like 256 bit, 256 different chips. And when it was just a Philips standard, that was great. Right. And if you don't know where it comes from, do you know? I don't know.

Chris Gammell: I mean, I could read through the wiki page further, but it's probably faster to ask you.

Dave Jones: Well, I'll tell you. Okay. Well, except for the long, detailed historical description that I'll give. Okay. It comes from, Philips used to design, they were the leading supplier of chips for TV sets. Uh-huh. And back in the day, you know, like 70s and 80s, TV sets were the leading consumer gadget. Now it's bloody mobile phones, right? Right, of course. But back then, right, it was TVs were driving, you know, the sales of everything and the components and all that sort of stuff. So Philips, they cornered the market. If you go look at any of the data books from the 70s or 80s, Philips have like a TV chip data book and just all the different chips for doing all the functions in a modern, quote marks, TV.

Chris Gammell: So CRT level, like color, multi-scan, whatever.

Dave Jones: Yep. They do the decoders and the, you know, everything, right? Yeah. Tuners, the whole works.

Chris Gammell: Quick aside, I was watching a Slow Mo's guys video the other day. They actually do, which is a great YouTube channel, if people don't know. Yeah. They did, they slowed down watching a television scan rate. Oh, yes. I will link it in. It is so good.

Dave Jones: Yeah. That's great. Yeah. And, yeah, so they were, you know, they had hundreds of these chips, right? And maybe not hundreds, but they had like dozens and dozens of these, you know, full data book worth of these bloody chips. And they, you know, and functions were getting more advanced that, hey, we want to tie these together. These chips are getting smarter. We want, you know, a push button TV, you know, or remote control push button, fancy digitally type TVs. And we need them to, you know, we need these chips to switch digitally instead of having the old school knob, you know, that clunk, clunk, you know, when you change channels.

Dave Jones: I know that one. You're right. Okay. I was there.

Chris Gammell: Maybe some of our younger listeners won't be.

Dave Jones: And these big multi-gang switches, which should handle everything for you. Right. So they went, yeah, we need a digital standard. And that's why they developed I2C. It's the inter-IC communication system. That's what I2C stands for. Oh.

Chris Gammell: I-C being integrated circuit, right? Inter-integrated circuit. Between integrated circuits. I got it.

Dave Jones: In between, and that's how they talk. But they couldn't, you know, in a TV, you can't just.

Chris Gammell: I mean, that's a name that came up with, that was an engineer that came up with that name. Inter-integrated circuit. Circuit. That would never happen these days. It would be like the synergy bus. Oh, right. Yeah, exactly.

Dave Jones: Yep, totally.

Chris Gammell: Yeah.

Dave Jones: I'll pay that.

Chris Gammell: Yeah.

Dave Jones: And they went, and of course, you know, TV's back. Like, you didn't have your four, six-layer boards. None of that rubbish. So you can't just route things willy-nilly, right? So they figure, yeah, let's do this multi-drop thing. So they developed this multi-drop thing that only required two wires. And that was groundbreaking. So you could connect to 20 different, you know, a modern TV might have 20 different chips in it. Right. And you've got to join them all up. Right.

Chris Gammell: Instead of having a wire harness that has-

Dave Jones: Yeah, yeah.

Chris Gammell: You know, two to the tenth wire. Yeah, exactly. But yeah.

Dave Jones: Yep. Exactly. So, and it was, yeah, it was very clever. And they originally designed it, hey, look, we just need to tie our own chips. We've got an 8-bit address system, 256 chips. That's plenty. You know? Right, right.

Chris Gammell: Who could ever need more memory, right? 600k, 80k, right?

Dave Jones: 640k.

Chris Gammell: Oh, 640, right. Yep.

Dave Jones: Yep. Famous Bill quotes. Yeah. Bill Gates quote. And they, and then other manufacturers went, oh, shit, we'll copy that. That's a good idea. And so they came out with these clone things. And then all of a sudden, well, 8-bit address system wasn't enough. And so, and Philips, you know, once they started, oh, we can make money from licensing this thing. It's like, yeah, they were running out of, you know, addresses to allocate to people. So they would give you a unique address. And then it got to the point where very quickly that simply they ran out of addresses. Then they had to switch to an advanced 10-bit addressing system. And then they still ran out. And then it got to the point where a lot of the chips on the market would share the address.

Chris Gammell: Yeah.

Dave Jones: So a lot of, that's why a lot of I2C chips these days still, like I2C memory, if you go look at like an 8-pin, you know, E2 Prom or whatever, that's I2C. It'll have three address pins on it, for example, so that you can set the address. Whereas it used to be hard-coded.

Chris Gammell: Right.

Dave Jones: Right.

Chris Gammell: Yep.

Dave Jones: You know, because Philips would give you, would allocate you that number, you know, and nobody else on the market, in theory, could use that number. You know, that was your address.

Chris Gammell: Which is interesting because that's also how the Bluetooth standard works. That's also how USB standard works. Like to get that, a truly unique address. Yep. That's really, you need some kind of centralized authority in order to do that. Of course, these days, Dave and I are launching blockchain synergy bus.

Dave Jones: Yeah. The blockchain hardware.

Chris Gammell: Should have seen that one coming from a mile away. Right. Yep. Yeah.

Dave Jones: We actually just got a UID. Is it a unique identifier for the USB bus? Oh, nice. For the EU blog meter? Okay. No, for the micro supply that we're working with.

Chris Gammell: Oh, not the meter supply. Okay. No. Okay.

Dave Jones: Yep. Yep. The micro supply. Cool. That's great. Okay. We got officially allocated, you know, so we can have one of those signed drivers, you know. Oh, yeah. That's great. You know, those signed Windows drivers. Yeah. So, you know, yeah, it's officially signed by the EV blog. We got our own number. That's cool, man.

Chris Gammell: Yeah, it's pretty cool. Did it cost money or what?

Dave Jones: No, we got it free. It's actually, we were surprised by that. You can actually get it free. I think we got it through one of the UART manufacturers. They offer that service. If you register with them, they will use their pool, their allocated pool. Oh, they have like a block of them? Yeah. They've got like a block or something. That's a good service though, yeah. You know? Yeah, it's great.

Chris Gammell: And you don't have to like, I guess you just have to keep buying chips through them in order to keep it. Oh, probably. Yeah. Yeah. Maybe. But at this point, you're probably doing that anyways.

Dave Jones: Yeah, probably pretty much locked into it. Yeah.

Chris Gammell: Right.

Dave Jones: So, anyway, so the I2C system, the interIC system. That's nice.

Chris Gammell: Yeah, I'm looking at the wiki page too, like the SM bus looks like that was the derivative of it by Intel.

Dave Jones: Yeah, there is a, yep.

Chris Gammell: In the 90s, so.

Dave Jones: Yeah, there's a couple of clones of it.

Chris Gammell: Yeah. Yep. It's interesting too because like, so like Bluetooth and USB and all these other things and I2C as well, I guess. Because this is like why there's so many standards at the beginning normally, right? They're all like competing standards because it's like, ultimately, they're trying to get mind shared, right? They're trying to get it to the point where it's like, they don't want to give up the trademark. They don't want to give up anything else, but they do want to be the standard, right? They want to be like the one that other people have to register with.

Speaker ?: Of course I do.

Chris Gammell: Everyone does. Yeah. So, that's why whenever you see multiple standards, you know there's money lurking behind the scenes somewhere. Right. Yeah.

Dave Jones: Boy.

Chris Gammell: Yep. So, yes. So, to get back to the troubleshooting at hand. So, what was happening? And it was really stupid at the end of the day. So, I got the, so I finally got the micro, the ESP32 outputting I2C code. No problem there. Was verifying this on a scope. And I was like, oh, I see the traffic. I know I have the data. This looks like data. This looks like clock. I have it plugged in, right? And I was just monitoring with the scope. Oh, did you see my floating scope setup? Did I show you that picture? Did I show you that picture?

Dave Jones: No. Floating scope? Yeah. Are you lifting the earth?

Chris Gammell: No. No, no, no. Wait. Let me send you a picture so we can get the real, the real reaction. All right. And I will post this picture.

Dave Jones: All right. A live reaction. Yes. Live reaction.

Chris Gammell: What the hell is it? Oh, here we go. All right. Sending to Dave now. So, floating scope. And that's, so I was just using the scope. Well, I.

Dave Jones: All right. I was going to say, is it floating in the air? And yes, it is.

Chris Gammell: It is. Yeah. It's, yeah. It's actually, I'm surprised I never thought to do this though. Like, so basically what I did is.

Dave Jones: What are the cable tires?

Chris Gammell: Yeah. I just zip tied the handle of the scope underneath. Like, so I have this like metal rack. Oh, sorry.

Dave Jones: I thought, I know which scope. But yeah, it's the Tektronix 3000. Yeah. The MDL 3000.

Chris Gammell: Yeah. All right. Yeah. And so, but I have this, this cable. So, I'm pretty proud of this little setup here. Basically, I took like one of those metal kitchen racks. Right. Yeah. And it's, it's actually meant for like a, like a, you know, three foot high. There's just one that you put in your kitchen somewhere. And I moved all the bench stuff up and then I mount stuff underneath it. Right. So, it kind of made like a little hood from my, from my open desk here. And I'll send, there will be a picture, like I said. Cool. But, you know, I always kind of struggle with having everything above the bench. Like, I know that your bench, you have your shelf. Yep. But I think that it just, you know, stacking stuff too high and it kind of gets really away from you. And so, I was like, oh, I have the scope right here. So, you know, hung it underneath. I thought it was pretty clever. It is.

Dave Jones: That's great.

Chris Gammell: Anyways, I was using that scope and I wasn't using like the protocol analyzer on it. I was just using the actual scope just to see the analog signal. Signals look clean. The edges look clean. That was all great. I made sure I had the clock going in the right place, the data going in the right place. You can, you know, tell if the clock always looks like it's very regular. The data has a lot of variation there. And so, I was like, ah, I can't figure out what this is. And then finally, someone's just like, well, you are looking at that on a logic analyzer, right? And I was not. Oh. No.

Chris Gammell: No. Okay.

Dave Jones: Right. Yeah.

Chris Gammell: Which was stupid, right? So.

Dave Jones: No. It's not stupid to look at the analog waveform. That's what you should be doing. Well, yeah, maybe. You don't use an analog. You don't use the logic analyzer off the bat.

Chris Gammell: Oh, that's interesting. Okay.

Dave Jones: No. Always look at signal integrity matters.

Chris Gammell: Interesting. Okay.

Dave Jones: That's a T-shirt.

Chris Gammell: I mean, Dave, you were talking to me. That is my MO as well because I think that's analog, right? But once I did that, I'm like, oh, it's fine, right? I guess my misstep was I said, oh, that signal looks like the signal it's supposed to be. I didn't actually check what it's actually saying, though. Because ultimately, what was happening is that. So, I did finally put on a logic analyzer. I have that analog discovery, too. I plugged that in. And what was happening was it just kept retrying the same address because it wasn't getting any knowledge. And that was a real problem.

Dave Jones: That's the classic I2C, NOAC. Exactly.

Chris Gammell: So, I had not really dealt with NAC before. Yeah.

Dave Jones: Well, yeah. It eventually got down to a protocol thing. Yes. Once you check your first step is to check your signal integrity, if that looks okay, and you haven't swapped the lines and done anything stupid, and there's data happening there, then you've got no choice. You have to go in and protocol analyze it.

Chris Gammell: Yeah. Yep. Yeah. And it turned out that it was actually the first board that I was working with. I must have lifted a trace or something because I just can't talk to the chip on that board. And so, I...

Dave Jones: So, you were talking... The master was talking to the slave, and the slave wasn't responding. Yeah.

Chris Gammell: I was just saying, nope, nothing here. Well, and really, it's not nothing. I guess the response is just acknowledge or not, right? It yanks down. So, to use that bus analogy like we were talking about before, right? If the address is correct, then... So, if I said address hex 77, then the chip goes... It pulls down on the line. It says, yep, that's me.

Dave Jones: It pulls down the line and say, yep, I heard you. Yes. Right. Please send me more.

Chris Gammell: And it was not there. Right. And, yeah. So, finally got to the point where I re-soldered another board. I had thought that the new board was not working either. And then, finally, yeah, I finally saw an acknowledge. And it was... Right. And then it starts piping data like it's no problem, you know? And that's the thing. Like, most people, when it just works, it just works, right? Yeah, yeah, yeah. Like, until you get... Like, I should... I don't know. I don't know what the hell my problem is. I feel like every time I...

Dave Jones: I2C is more complicated than, you know. It's, yeah. There is more things to go wrong.

Chris Gammell: Yeah, but it's just, like, the chip on the other end, the slave chip, is so stupid. You know what I mean? It's basically, like... Yeah. It's like a set of registers hanging off a bus.

Dave Jones: These are state machines, yeah. Yeah, exactly. You know, these aren't intelligent controllers. No, they're just dumbass state machines. Yeah.

Chris Gammell: Well, that's what's super frustrating, though, is that, like, I'm like, it should just be working. You know what I mean? Like... Yeah, right. This is, like, the Dave Jones-ism, right? Of, like, I hope your project doesn't work. And it's like, yes, I learned a bunch. But I didn't want to. You're welcome. Damn it, Dave. Like...

Dave Jones: Sorry. Yeah, just when you have to get shit done, you know. It's like... Yeah, exactly. Oh, no. Not...

Chris Gammell: Right, right. Learning's meant to be when I have time, and it's never... Yeah, right. Yeah. Yeah.

Dave Jones: I love it. I didn't finish telling the story.

Chris Gammell: Oh, yeah, yeah, yeah. The driver.

Dave Jones: Of my... Of what... Yeah, I was developing a product.

Chris Gammell: Mm-hmm.

Dave Jones: Right? I got this idea that I'd corner the market. Yeah. I'd corner the market for an I-squared-C tester that you could just whack on a bus, right? And it had, like, a little, you know, two-line... As you called it at the time, like, a little two-line, you know, LCD and everything. And it decoded the I-squared-C bus. It told you what's going on there. Faults. It was designed for, like, TV repair people. Yeah, right, right, right. You could just whack it on the bus and go, well, is this bus working? Is it doing shit? Is, you know... Are all these chips talking, you know? Right. And then you can go, aha, that decoder chip over there is not talking. Replace that chip. Right. It's like, oh, look around there, there's something, you know, something to do with that chip.

Chris Gammell: And this is kind of what the... I mean, like, obviously, there's a lot of stuff that does this, but the single-use type of thing, that's kind of what the bus pirate does, too, right?

Dave Jones: Yeah, but it tries to be more universal. It does. You're right. Mine was going to be, like, a more hard-coded, specific thing.

Chris Gammell: How are you thinking through, like, the actual, like, stepping back? It would just record all the messages that it saw, and then you'd kind of step back through them?

Dave Jones: It will. And it would... No, it would analyze it. It would smart analyze it for you, and it'd have a database of things that, no... Okay, it could tell you that there's five chips on that bus, you know? Oh, I see. So it scans through all the addresses? Yeah, yeah. It scans through all the... Like, it does everything, you know? Okay. Yeah. And, yeah, so I thought, you know, that'd be useful. And then I realized the amount of work on the software side.

Chris Gammell: That would be very software-heavy, yep.

Dave Jones: But, you know, I wrote my own I2C driver, and, you know, and, you know, it was working. It was doing the business. But, yeah, it was just, like, a lot of... Yeah. It was just one of those never-ending projects software-wise, you know?

Chris Gammell: Yep, yep.

Dave Jones: Whereas the bus... Maybe, you know, it would have ended up something like the Bus Pirate, you know?

Chris Gammell: Yeah, yeah. I mean, yeah, Bus Pirate's great for, like... It's a tool. You're never going to get rich making a tool, right? But it's... No, no, that's right. But it's pretty universally loved. I know a lot of people that are, you know, they love the Bus Pirate, so...

Dave Jones: But back in the day, you know, like, there were X number of TV repair tech, you know? Yeah, of course. There were, you know, 50,000 TV repair techs, you know? If you could sell one to every TV repair tech, you know, that's a pretty good business, right?

Chris Gammell: Well, I even remember, like, when you started reviewing, like, the Rigols at the beginning, like, they didn't have protocol stuff included, like, these days. Oh, no, no. Every cheap scope has a protocol analyzer in it now. Exactly. It's totally Chinese. Right. And that is a true blessing, though. Like, I mean, I didn't use it at first, but, like, to actually have that there... So if people are out there and they are, you know, like, staring at their CRT screen and trying to decode messages themselves, it's a good exercise at first, but it gets real old real fast, you know? Yeah. And, like, so it's good to, like, know the shapes that you're looking at, obviously, but being able to just see, like, the blocks of addresses, because it is. It's like, you see, you know, eight bits going up and down, right? If it's... Well, you see the clock going up and down eight times, and then you see the address, you see whatever you need to see there, and that really... Yeah, that really helps. Yeah.

Speaker ?: Yeah.

Dave Jones: But, yeah, I mean, this was the days before there were, you know, these cheap USB logic analyzers that are almost free, you know? You can get them 10 bucks on eBay for these rip-off logic analyzers, you know? Yep. And, yep. Different world.

Chris Gammell: Different world, man. Cheap logic. Cheap logic. Did you end up listening to the FPGA episode last week?

Dave Jones: No, I didn't. Sorry.

Chris Gammell: I was like, Mr. Luke Valenti of Tiny FPGA. It was very good. I think that's the thing, like, so, like, if people are looking for a project, like, you could go and take something like that Tiny FPGA project and just go and make a decoder, right? Or make an interface. You're right. You know, like, FPGAs are great for that. So, like, this Analog Discovery 2 I have is basically an FPGA with front end on it. You know, that's a lot of what the low-cost tools are, either micros or, you know, FPGAs to just make it configurable and fast and handles logic well. So, yep.

Dave Jones: Cool bananas. Anyway, so I didn't end up cornering the TV repair market.

Chris Gammell: You were not an I2C maven, you're saying, yeah?

Dave Jones: Could have been. Coulda, shoulda, you know? Yeah. Yeah, being filthy rich now. Yep.

Chris Gammell: Yep.

Dave Jones: Oh, dearity. Yep. Yeah. Boy. You're right. What else have we got? Oh, speaking of, like, well, no, it's not. But I like this. I like this chip. It's chip of the week. Oh, sure. Can we have a chip of the week?

Chris Gammell: Yeah, yeah, yeah.

Dave Jones: Yep. Chip of the week from Diodes Inc., which you don't normally think of chip of the week would come from.

Chris Gammell: Usually not super complicated parts from Diodes Inc., but, yeah.

Dave Jones: Anyway, the PT8A2511, huh, know it well. It's a toaster controller.

Chris Gammell: From your toaster days, huh? Yeah.

Dave Jones: I love the toaster days. I've got a story about the toaster days.

Chris Gammell: Yeah, the heady toaster days of the world, yep.

Dave Jones: Which I'm sure I've told before on the Amp Hour, but I will tell again in a minute. You have to remind me. Okay. And it's a toaster. It's exactly as advertised. It's an SO8 package, and it's designed to go into toasters. It's got a defrost mode for frozen bread. It's got a reheat mode.

Chris Gammell: It's a very specific pull-down pin, you know? Yeah.

Dave Jones: Yeah. There's like a reheat pin, you know? And there's a relay pin, and there's a like...

Chris Gammell: I mean, I think this is like a lot of other chips that are out there, though, too, where it's like this is a true application-specific IC, right? This is an ASIC. Like, even though it's not a super complicated one, it doesn't get much more specific than this.

Dave Jones: And it works at 1 kilohertz. It's got an oscillator pin. Don't know if it has a built-in oscillator. You can use an external RC oscillator. I don't know this crystal rubbish. Right, right.

Chris Gammell: If you want to be like super precise, you should do a video with this where you get this controller, and you put like your frequency standard into it.

Dave Jones: I put a rubidium oscillator into the toaster to time. Yep. Yeah. That's brilliant. Yep. If it's super accurate toast, it matters, you know? It does.

Chris Gammell: It does.

Dave Jones: And, you know, a pin which shorts to ground once an active negative pulse appears at defrost, you know, it's just... It's great. I love special purpose. Well, actually, it doesn't come from Diode. It's ink. It comes from Pericom.

Chris Gammell: Yeah.

Dave Jones: If you look at the data sheet... They probably bought them at some point. They bought them. Yeah. Anyway, Pericom looks like they own the market. For toaster controllers. Yeah. You know? I wonder how much this thing costs in volume, you know?

Chris Gammell: I'm trying to find the tweet. Someone tweeted at me. They actually bought one and then already implemented it. Right. On... It was just a breadboard thing. Yep. They just out of the chip and, yeah, it worked. So, it's great.

Dave Jones: Oh, I love it. And, like, they've got a typical application schematic and none of this linear regulator rubbish in there. They've got a 3.9 volt Zener.

Chris Gammell: Oh, yeah.

Dave Jones: Because Zener's are cheap, you know? Yeah, you go cheaper. Right. We can go cheaper. Yep.

Chris Gammell: Yeah.

Dave Jones: Oh, man. Yep. I'm surprised they bothered to put a back EMF diode on their relay coil. Jeez, that costs a couple of cents extra.

Chris Gammell: I know, right? This is muncing territory, huh?

Dave Jones: Yeah. Fancy, fancy. And they're, like, they're literally, like, tapping. They use, like, the 3.9 volt Zener comes directly from the main. So, what they're doing is taking it through the heater, right? So, you've got 240 volts in, right? Or 110 or whatever it is, right? It goes through the heater coil, right? So, it's got the lever. It goes through the heater coil and then goes into a resistor divider. They tap off the resistor divider into a single-wave diode rectifier just to rectify the AC mains, right? Yeah, I can't afford that full-wave rubbish. No, no, of course not. It costs a fortune, right? Right? So, half-wave. Sorry, that's a single-wave. Half-wave is what I'm thinking of.

Chris Gammell: It's a single-wave if the wave's big enough, you know? Yeah.

Dave Jones: Half-wave rectified. And then it goes into a 3.9 volt Zener to power the chip and actually provide the pull-up for the leads and everything else. It's just great. That is awesome. Yep.

Chris Gammell: It's crazy, too, is that even the $80 toasters that are out there have things like this in it. Oh, yeah. Because why wouldn't they, right? How much more complicated does it need to be?

Dave Jones: They've got 30 cents worth of parts. Right.

Chris Gammell: I mean, the fact is, this is putting a piece of silicon into a toaster, which it could do the same thing with, like, what's that metal strip that's heat-sensitive?

Dave Jones: Yes, the bi-metallic strip thing, which, yeah, old-school ones probably use.

Chris Gammell: That's what the $6 toasters are, right? Right. Because in that $6 toaster, it's like, how do you get something from China to the United States and sell it with margin for $6? And it's like, there's no electronics in there.

Dave Jones: But this one, this is, I just love this, this special-purpose chip. Yeah. It's just, it's great.

Chris Gammell: So, I did find Edward Shin had posted about this. I sent the link to Dave. Dave, he did, he did try it out. And, uh... Sweet. Yep.

Dave Jones: And it works.

Chris Gammell: It works, yeah.

Dave Jones: In our DIP-8 package, none of this SO rubbish. I love it.

Chris Gammell: Yeah, it's great. He needs to power cycle to toast again, apparently. But that makes sense, right? Because that's actually what the lever on a toaster is.

Dave Jones: Oh, okay. So, he's a firmware engineer at Fitbit.

Chris Gammell: Yeah.

Dave Jones: I thought Fitbit went bust. No, they got bought by something. No, not Fitbit.

Chris Gammell: Pebble got bought by Fitbit, so... Right. Yeah. Yeah, Fitbit's still around.

Dave Jones: Okay. Yeah. That's just great.

Chris Gammell: Yeah. Well, all you need to know is, you know, get an ESP-8266, hook it to this thing, boom. Internet connected toaster.

Dave Jones: Oh, yes. Internet connected toaster. That's all the bloody world needs.

Chris Gammell: That's what you need, man. You never... You never know.

Dave Jones: Ugh. So, can I retell my toaster story?

Chris Gammell: Oh, of course, yeah. I want to hear that.

Dave Jones: Right. I can't remember what year this was. It was maybe mid-90s or something. And, you know, I was looking around for a job, as I always did, you know. I was always looking, always reading the paper.

Chris Gammell: Uh-huh.

Dave Jones: Back then, right? The paper.

Chris Gammell: Circling the potential ads with a pen.

Dave Jones: Exactly. With a pen, you know, a highlighter, you know. Yeah. I'll call him up, you know. And, yeah. And there was this job at, like, it was nearby, you know. And it was like a design consulting company. Oh, okay. Yep. You know, yeah. So, I thought, oh, yeah, cool. You know, like, oh, yeah. Oh, sounds good. You know, it's local. And so, I went in to meet the guy. And he was like, oh, man. He was... God, he had to be 80, I think. Like, he was... He was an old fella.

Chris Gammell: He had glasses. He had glasses. You knew he was old because he had glasses. Sorry. Couldn't resist. Couldn't resist. Nice callback. Nice callback.

Dave Jones: I'll pay that. All right. And, anyway, it doesn't matter. I just noted that he was old. Anyway, and one of the jobs was that they had this contract to design this newfangled toaster. Oh, really?

Speaker ?: Yeah.

Dave Jones: From this company, right? This company had this patent on, like, instead of having elements, right, that they would have, like, heated plates. And the plates would come in and touch the side of the bread. And they... Oh, wow. But that wasn't what their patent was for. I think their patent was for the way they cooked the toast was that they would cook it at a lower temperature for longer. And I... Low and slow, like barbecue. Low and slow, in direct context to the plates, and they could control the temperature profile, just like a thermal oven. You know, they could control the temperature profile of the toast. And apparently, like, it comes out perfectly golden every time. Wow. You know, it was, like, perfectly consistent, like, you know.

Chris Gammell: I'm always toasting bread, and I'm like, you know what would be great? Yeah. Is if this was slower.

Dave Jones: Right. Yeah, exactly. I know. But you get your perfect toast, you know.

Chris Gammell: I guess, yeah.

Dave Jones: Yeah. Anyway.

Chris Gammell: I have a similar, a better... This is, like, a better mousetrap, better toaster. Right. Currently at M-Hub... I think I'm allowed to talk to this, because they work on this out in the open. Currently at M-Hub, the place I work out of, there is a group of people working there using machine vision to toast bread and to get the perfect toasting. So, this is... Yeah. 20 years later, this is... The only difference is just the newest technology. Right.

Dave Jones: Right. Oh, my God. And that's really what you want as a camera near a heating element. Oh, yeah.

Chris Gammell: Yeah, that's what you want. Right, right, right. I mean, I know some of the people working on it. And, like, they're doing this job for, you know, it's a client, whatever. You're right. Yeah, yeah, yeah. It's a gig. Gig's a gig, you know.

Dave Jones: Yeah, they've probably got millions in VC funding for this bloody widget, right?

Chris Gammell: No, definitely. I don't think so. But the problem for me is that I walk in and I'm like, it smells like toast in here. It smells delicious. Like, they're just toasting all day long.

Dave Jones: All day long. Yeah. And it's... But you've got to have millions. If there's a team of contractors working on it, you've got to, you know... It's a unique research project.

Chris Gammell: I think that's all I can say about it. Yeah.

Dave Jones: Anyway, yeah, it's cool. Anyway, so I'm deciding whether or not I should take this job. You know, they had some other stuff. But I would be the primary design guy for this toaster.

Chris Gammell: You would be the toast master? Is that right?

Dave Jones: Yeah, I would be the toast master. And back then, this was before the EVE blog forum.

Chris Gammell: Of course. Right?

Dave Jones: Yeah. And I won't name the forum, but I was on a forum that... An engineering forum.

Chris Gammell: We all know what it is. Usenet. Yep. It was back in the Usenet days. Side tech, dead electronics, whatever it was back in the day.

Dave Jones: I won't name the forum. Anyway, there is a Usenet group. And there was a famous troll on there. And he repaired... One of his jobs was repairing toasters. And I won't name him. And everyone made fun... Like, they'd call him toaster boy. Oh, man. And it was just like the running joke. Yeah. Like, you know. And it was like so... Like, yeah, because this is the community I hung out in. Right. And if people found out I was designing toasters...

Chris Gammell: Right. You would have been the new toaster boy.

Dave Jones: I wouldn't have been able to hold my head up high in there. Like, because it was like, you know, it was the, like, joke that, oh, he's so crap, he's repairing toasters for me. Right, right, right, right. Like, you know, it was one of those, yeah, right. And it was one of those derogatory, you know, things. And it's like, yeah, I'd be the world's best toaster designer, you know. And so I went, no, thanks. Oh, my God. So that was the decision to actually turn down that job. Oh, that's specifically why. Because I was actually offered the job. Right, right. Yeah, yeah, yeah. That's specifically why, yeah. Wow. Wow. Yeah. That's great. I didn't want to be branded for this toaster boy.

Chris Gammell: Wow, Dave Jones making decisions based on strangers on the internet for decades now.

Dave Jones: Yeah, totally, totally. Yeah, I bet that that was my community back then, you know. Of course, of course.

Chris Gammell: I mean, yeah, that's like a lifeline, right? And that's great. Yeah, yeah, totally. That's what's morphed into the EV blog. Well, hopefully less trolly, but I've been on the EV blog forum. It's a little trolly. Right, well, you know, we're trying. It's like the most helpful trolls out there. Right, yeah, totally. We're talking to head honcho troll right now, so.

Dave Jones: Anyway, toaster. Oh, that's great. It's just funny to hear that they're still doing it. Yeah. With the vision, machine vision. I know, right. Oh, man. And you just know that shit's going on Kickstarter, you know.

Chris Gammell: Yeah, we'll see. We'll see. Yeah. Yeah.

Dave Jones: Oh, I posted. Sorry, I don't get the link, but I tweeted the other day, yesterday or the day before, I, once again, it's not new news. It was like from last, late last month or something, so it's a few weeks old, that this Kickstarter failed. It's for a smart headphone. What is it? The Oc...

Chris Gammell: Ocic, yeah.

Dave Jones: Ocic, Ocic. The smart headphones.

Chris Gammell: You tweeted about it about six days after I put it on the Amp Hour subreddit. Oh, you did. Oh, okay. I wonder where Dave heard about that. Yeah, that was an unfortunate shutdown, especially given how much frigging money they took.

Dave Jones: Six million bucks. It's crazy. They took three million dollars of backers money, plus they had another three million venture capital.

Chris Gammell: It is with an extremely heavy heart that we must inform you that Ocic is shutting down and will be unable to deliver the remaining headphones.

Dave Jones: Anyway, apparently they're like super whiz-bang 3D. Yeah, it was meant for like AR and VR type things. Virtual reality and all VR sort of stuff. Yeah, yeah. You know, and so a lot of people started attacking them on the forum and I said, hey, you know, look, the problem is, because we talked to Jerry, right, recently about this whole thing. You know, once the VC people come in, right, it's like your company changes focus, right?

Chris Gammell: I don't think so. What?

Dave Jones: Jerry's company totally changed focus once they came in.

Chris Gammell: I agree with that. But if people have not listened to Jerry's thing, I recommend that they go and talk to him. But here's the part. And maybe this is, you know, an argument against Jerry too, but inventing something, so this is a quote that I quoted in the comments here. Inventing something new while also developing complex hardware is expensive. The addition of stretch goals to add mobile support increased the software scope from two operating systems to five, added an incredibly powerful 32-core processor onboard for headphones for processing, and requires us to enter into substantial business development with mobile manufacturers to support multi-channel connectivity. It ultimately doubled the size of our development. Basically, that's scope creep. It's scope creep. Like, Jerry was doing the management stuff. This is pure scope creep.

Dave Jones: Well, it's not pure scope creep, because once that VC money comes in, right, they go, right, we're going to put a CEO in. We want this to be a real business. We want to have long-term goals. We want a long-term exit strategy. We don't care. You know, like this Kickstarter thing. Okay, fine. You know, you can still play with your little toy, right? But we want this to be a real business. And it kind of, you know, it can change the focus of the company rather than, oh, look, we're a hungry little startup, you know, and we're just focused on delivering the world's best headphones. And now suddenly, oh, you've got a new CEO and a new direction. I guess so. And people tell you you've got to become a real company. You know?

Chris Gammell: I mean, it's $6 million though, Dave. Like this is, Jerry and Rick took a million and lasted for a long time versus this. So I'm just thinking this sounds like this was a experience thing personally. Again, I haven't looked into it enough though. So maybe I'm being the trolley old guy now.

Dave Jones: That's why I also think it's probably a lot of it has to do with the VC people like said, no, like we want, you know, we're going to hire all these people and we're going to turn this into a real company. And bingo, the cash burn rate just goes through the roof.

Chris Gammell: That is, you know, if that was the problem, then yeah. But yeah.

Dave Jones: I mean, you've got to burn through the cash somehow, right? Because I'm assuming that they didn't skip the country with the money, right? I'm assuming that they simply ran out of money and time and they just burnt it all, right? And that's one of the ways to do it is to, you know, hire everyone.

Chris Gammell: How do you even fix this in the first place, right? Like they obviously promised things that they were not capable of doing if they switched if they switched it around to, to a, from a two, you know, going from two processors to five and all these other things, obviously it was not.

Dave Jones: And I agree. Stretch goals are stupid, right? Because if you're adding a stretch goal, it's a last minute thing.

Chris Gammell: It's like designing in feature creep, right? Yeah.

Dave Jones: Because you, you, oh, my Kickstarter might be successful. Oh, I better add a stretch goal. And, and then you don't do the proper engineering behind that and go, well, is this a good idea? What's it going to do to my, you know, development cycle, blah, blah, blah, you know, the bottom line. And you just don't analyze it. You just throw it in there as, you know, because that's what you do on Kickstarter. You know, you have stretch goals once you get popular, you know, to try and make it more popular and, and they just don't analyze and think about these things and you can come a gutzer.

Chris Gammell: Yep.

Dave Jones: Right.

Chris Gammell: Well, it's a damn shame, but a reminder, if people are backing Kickstarters, sometimes, sometimes they go belly up. That sucks.

Dave Jones: Well, see, if I, if I was running that, if like, and I promised all these stretch goals and I suddenly realized, hey, this is, you know, shit, we're, okay, we're targeting five platforms instead of two or whatever it was you said, you know, then I would eventually go, right, sorry, we're going to have to fess up and go, sorry. Some of these are, yeah, right. We're just doing iOS. We cannot deliver the stretch goals. Right.

Chris Gammell: They, they delivered, that's going to be the most money at the beginning. Right.

Dave Jones: Yep.

Chris Gammell: Yeah. Well, I have not done this, so I can't speak any further on it, but it sounds like this is unfortunate.

Dave Jones: Apparently they did ship some units, but yeah, it's one of those unfortunate things that they just, they burn through the cash. And, and we've talked about this before. You can get, you know, the, the valley of death, isn't it?

Chris Gammell: You know? What's that?

Dave Jones: Where you, huh?

Chris Gammell: Well, which, which, which valley of death?

Dave Jones: Oh, well, the, the income valley of death. You, you, you get.

Chris Gammell: Oh, this is the SparkFun post of, yes.

Dave Jones: Oh, I can't remember the exact where it came from. But yeah, like the, you can get, there's a point where you can get not too much money, but kind of like a lot of money, but not enough. Right. Right. And then it, yeah.

Chris Gammell: What do they call it? Crowdfunding something.

Dave Jones: I'm sure it's something like that. We've talked about it before. It was a.

Speaker ?: It was a. It was a. It was a. It was a. It was a.

Dave Jones: It was a.

Chris Gammell: It was a. Yeah.

Dave Jones: Yeah. Yeah. It, you know, you get, you get, you get wildly more successful than you thought, but not wildly, wildly successful. Right. In order for it to be a problem. You know? Like. So, yeah. There's that middle ground there where it, where it becomes a problem. Yeah.

Dave Jones: Yeah. Yeah. Yeah.

Chris Gammell: I'm trying to find it furiously. I can't find it. So. Oh, well. Anyway. There is a post out there somewhere.

Dave Jones: Yeah.

Chris Gammell: Yeah. Well, that's a shame.

Dave Jones: Yeah. Anyway. Yep. Oops. Yeah.

Chris Gammell: Boy, we got some other stuff on here. There's a great post by Tom Anderson, who's a. So Tom writes for Tempo quite often. Yep. But he also is a longtime oscilloscope nerd. And, but this is just like, he, he also like does his own animation. So he wrote, he wrote about latest USB-C and it's definitely a great resource post. So I'm going to link that into basically animations of the connectors themselves and how they mounted the board and all the cabling and stuff you can do with them. And then the speeds you can get. So just a great post by Tom. So very cool. Um, very cool resource for people.

Dave Jones: Excellent.

Chris Gammell: What else is on here? Oh, oh, uh, past guest, uh, Ken Schuriff.

Dave Jones: Oh, yeah. He's just killing it.

Chris Gammell: He's killing it. The 76477 Space Invaders sound effect chip.

Dave Jones: From 1978.

Chris Gammell: I have no experience with that, but, uh.

Dave Jones: Oh, yeah. This is great.

Chris Gammell: So basically it's like a little, it's like a synth, right? It just has to be like a mini synth to do all the different, uh, white noise and stuff like, yeah, all that. Yeah. Yeah. And it's probably not many layers either, but basically, uh, Ken goes through and he actually does the die shots and.

Dave Jones: Yeah. And it's got noise, um, generators and it's got attack and, and decay system blocks and, you know, things like that, you know, local oscillators build on it. Oh, you include the block diagram too. Okay. Yeah. So basically you can. Yeah.

Chris Gammell: So what do you, do you actually have to inject, uh, analog signals or digital?

Dave Jones: No, no, none is digital rubbish. What are you talking about? No. No. No. No. No. No.

Chris Gammell: How do you actually control this thing then?

Dave Jones: This is analog.

Chris Gammell: So this is an analog chip, but what do you, no, you inject a random noise output from things. So what do you just, you just knock it high or low and you say, start now, stop now. Yeah. It's yeah.

Dave Jones: That's it.

Chris Gammell: Got it.

Speaker ?: Yeah.

Dave Jones: Yeah. Oh wow.

Chris Gammell: Look at these waveforms too. Okay. Well, we'll link, we'll link it in, uh, it is in depth posts that I obviously have not read.

Dave Jones: So anyway, he's reverse engineering the die and, and he overlays block diagrams and then he draws all the die reverse engineers, the block diagram and just, it's just insane. I mean, the amount of detail and he's even got references at the end of his article. I want to know how long it takes to, it probably takes him a month to do that article, you know? Yeah.

Chris Gammell: Well, that's a shit ton of work. If you have not listened, definitely go back and listen to the, the episode with Ken. He's a, he's a bright dude.

Dave Jones: Fantastic.

Chris Gammell: Any other things we should add?

Dave Jones: That chip has its own Wikipedia page.

Chris Gammell: Does it really? Oh wow.

Dave Jones: Yeah. Fantastic.

Chris Gammell: Um. Sorry. Oh, I also posted, uh, so because I've been writing code again, oh good Lord, uh, I posted the Google style guide. Do you know that they, Google has a C++ style guide?

Dave Jones: Uh, probably doesn't, that doesn't really surprise me.

Chris Gammell: I, I didn't know they wrote C++, but I guess, you know, we've, I guess they do, right? I mean, I always assume that they do these high level, like D and whatever the go and whatever else they do there. Um, so I was just looking at like, I was asking around, actually I was asking the embedded folks, um, like my code, I'm always really bad at naming things when I'm writing code, like naming variables. It's just like thing one, thing two is like, you shouldn't do that, Chris. That's a bad idea. Yeah. So, so they had pointed me to this and it's like, this is very deep, right? So it's like, there's different formats for like, uh, uh, static members versus global functions versus local variables, like actually having different naming types around these things. And, and like, uh, so like, uh, a pointer always has a certain format and, and a thread always has a certain format and all these things. So it's very, like, it's very, very, uh, rigid in terms of this coding standard. But, um, yeah, it's, it's something, something to start from. I mean, see, I can develop my own, but it's, it's kind of, you have to pick something and stick with it. So.

Dave Jones: Hey, we, we totally missed a segue. What's that? In the, uh, blowing the budget, startup blowing the budget thing.

Chris Gammell: Which one?

Dave Jones: Apparently, here's the headline. It's apparently possible to blow through a hundred million bucks building a smartphone.

Chris Gammell: Oh yeah.

Dave Jones: And it's like, yeah, it's.

Chris Gammell: Right. Do you recognize the name of this too?

Dave Jones: The name?

Chris Gammell: Yeah. So. Of the phone? Yeah. Scroll down a little bit. For Essential, the startup founded by creator of Android, Andy Rubin, it was apparently a hundred million.

Dave Jones: Yeah.

Chris Gammell: Still not, still not ringing a bell? That's the guy that runs Playground Global, the, the accelerator that Jerry was based out of.

Dave Jones: Oh, right. Yes. Okay. Yes. Yes. Yes.

Chris Gammell: So, yeah. Um. Oops. Well, no, it's, it's fine. I mean, it was known. I mean, like there was, this was a known thing that like basically, even though, uh, even though they took on other people there and they were an investment fund, this was like funding the, one of the things the fund was funding was, was Andy's actual project, which is the Essential phone. And it's a gorgeous, I mean, like this thing is gorgeous, right? It's like an aluminum phone. It's. Well, it was. Right. Looks like an iPhone. Right. And, uh, but yeah, the idea was it's going to be this whole brand new thing, but like now it has, it's, it's nothing. Right. And so it's dead, is it? He's, he's trying to sell it basically. Oh, okay. Right. So, I mean, like the thing is Andy is going to keep doing this. Like Andy is a very talented person, right? I mean, he started Android and whatever, but once he got acquired by Android, if people don't know the Andy Rubin story as well, he got acquired by, Android got acquired by Google. He was at Google and then, you know, the guy who bought all the robot companies for Google was Andy Rubin. Right. And then now Google has all these robot companies, including Boston Dynamics and then Andy Rubin leaves and then he starts Playground Global and then he's doing this. Then he's also funding like, and you know, power to the guy, right? Like.

Dave Jones: Yeah. Yeah. But he, but he can't be too smart if he goes into the smartphone business without a phone that has anything that really stands out from the competition.

Chris Gammell: Well, I mean, if you've got hubris, you got one thing, you know, like I think that's the thing. Like it's like, but it's, it's fine. Right. I mean like that's, if there's money behind it, then maybe there was, there must've been some plan there that, that at least convinced the investors. Right. I'm sure that walking in a room and being like, yeah, I created Android is like a pretty big one. Right. It's like a proof point where they're like, oh, okay. But yeah, I don't know. Like making a phone is a, is a, is a, is a big deal. And I don't know what, what would even make it stand out these days? Like, so there were other things on it. The essential phone also had like, it had like plugins that could, you know, you could add like a camera to the back and.

Dave Jones: Yeah. I don't know. Make it just like give you like a two weeks battery life or something. You know, I don't know. Target like. Yeah. Why don't I just do that? Well, I don't like, you know, don't make it so right. You know, I have a requirement for a new smartphone. A, a screen that doesn't break and B, a battery that lasts like three or four days.

Chris Gammell: Okay.

Dave Jones: And that, that's the smartphone I went out and buy the smartphone that matched those, those two requirements for me. I went out and bought that one. Did you really?

Chris Gammell: How, how, how heavy is it?

Dave Jones: It's not that heavy. It's the Motorola Droid X or whatever. Yeah. And, and it's got a shadow, guaranteed shatterproof screen. I mean, they all say that. Five years.

Chris Gammell: Come on.

Dave Jones: Sorry.

Chris Gammell: They all say that.

Dave Jones: No, they don't.

Chris Gammell: Yeah. They all, I mean, they all say shatterproof. And Gorilla Glass and whatever, but it's like.

Dave Jones: Anyway.

Chris Gammell: Try, try dropping it without a case right on a corner. We'll see.

Dave Jones: Yeah. No, I have. Oh yeah.

Chris Gammell: On purpose?

Dave Jones: And it's good. Yeah. It's good. Wow. Well, not on purpose. I feel bad for your future glasses. And, and, and, and the battery lasts like three or four days in the damn thing.

Chris Gammell: Yeah.

Dave Jones: You know, like. That's great. That's exactly what I want.

Chris Gammell: Yeah. I know. Everything else just piss off. But you're not, you're not, you're not consumer A, you know, you are. Oh yeah. You are, you are not the bleeding edge.

Dave Jones: Yeah. Anyway.

Chris Gammell: Yeah.

Dave Jones: Hmm.

Chris Gammell: Well. I hope.

Dave Jones: My, my needs matter, damn it.

Chris Gammell: Yeah.

Dave Jones: Yeah. Whatever.

Chris Gammell: Uh, speaking of former guests, uh, former guests, Noah Feehan will be here in Chicago for a meetup that I'm doing. Uh, so if people are in Chicago and want to go to that, it'll be on June 12th. Friendly reminder, Dave is still not telling them another meetup in Sydney, but he's going to, I'm sure. Shut up. He's, he's super social. Just get David Tudor to do it, man. He's, he's a social guy.

Dave Jones: He's a social guy. Not really. He's not on social media.

Chris Gammell: That's true. That's a good point. Yeah. Anywho, if other people are, I was going to, the reason I actually wanted to bring this up, if people are coming through town, it's a great excuse for me to do another meetup. Um, and now there are four 3H meetups in the world. So. Cool. One in Houston, one in Austin, one here in Chicago.

Dave Jones: I'm not rebranding my one. Don't even ask.

Chris Gammell: That's fine. And then one in Seattle. Yeah. So there's four, uh, and it doesn't take much to brand yours. So if people want to start on the meetup, we will talk about it here. Cool. All right.

Dave Jones: We're blowing our budget for the show.

Chris Gammell: Yeah, it's an hour.

Dave Jones: Blowing our a hundred million micro amp hours or something.

Chris Gammell: Nice. All right. Well, there's other links on the subreddit if we missed anything.

Dave Jones: Yep. Cool bananas. All right, man.

Chris Gammell: Talk to you soon.

Dave Jones: Catch you next time.