#274 – Our First Call In Show

Dave Jones: This is the Amp Hour Podcast, recorded November 4th, 2015. Episode 274, the Amp Hour's first call-in show.

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: And it feels like episode one again.

Chris Gammell: It does. We're like nervous. We're giddy like schoolboys right now.

Dave Jones: We're just jumping in, both feet.

Chris Gammell: This is going to be terrible. We'll see.

Dave Jones: Kicking and screaming.

Chris Gammell: Yeah.

Dave Jones: We're attempting to do a live call-in show today, something completely different. Although something we've talked about since day one, I think.

Chris Gammell: Well, so actually it's been about one year since Tom Maglioli. I forget how to say that. Tom and Ray from the Car Talk. So Tom passed away about a year ago. And this wasn't really, it's just kind of coincidence. But yeah, we're trying a Car Talk style call-in show. But we're not airing it live.

Dave Jones: Except for the people who call in. If you call in live, you can actually listen to this. But there's no point telling you that now because you won't hear it until you know. You have to follow us on Twitter. If you're not following us on Twitter, you don't find out about this sort of stuff. So, yeah.

Chris Gammell: Right. And so I've had a couple other people send me the link of... Oh God, what's his name? The guy who plays Tobias Funke in Arrested Development. It's from Mr. Show. There was this spoof show about a pre-recorded call-in show. And it's like, people are always calling in from the week before. I'll post a link to it. It's really funny. Oh, right. Okay. But yeah, pre-recorded call-in shows are kind of redundant. But, you know, it's to get the... We don't want to be too pre-canned, you know. We don't want to prepare. God, no.

Dave Jones: Well, let's see how it works. It could be a complete and utter flop because nobody could call in. In that case, it's just us.

Chris Gammell: That will be the interesting point if anyone actually does call in.

Dave Jones: Or maybe Randy can call in. Oh my God.

Chris Gammell: Like, legit Randy. Not Dave as Randy.

Dave Jones: And on the line to help us out, we actually have a producer. Woo-hoo.

Chris Gammell: Yes, producer.

Dave Jones: The Amp Hour producer, Gary. Yes. Can we put Gary on? We can.

Chris Gammell: All right. Here we go.

Dave Jones: The magic of the interwebs.

Chris Gammell: Yes.

Dave Jones: Hey, Gary. How you doing?

Chris Gammell: Hi there.

Chris Gammell: I'm doing great. And you guys? Good.

Dave Jones: We're nervous as hell. Yeah.

Chris Gammell: Well, a little nervous.

Dave Jones: No. And tell us where you're from, Gary.

Chris Gammell: Okay. I'm from Paraguay in South America.

Dave Jones: Yeah. We've really got an international show. Paraguay, that's wedged between Brazil and Argentina. Is that right?

Chris Gammell: That's exactly right.

Dave Jones: Somewhere?

Chris Gammell: Yep. In the middle of South America.

Dave Jones: Oh, there you go. Middle of South America. Wow. Awesome. How's the internet there?

Chris Gammell: It's great. We have some connections. It's very expensive. But yeah, it's good.

Dave Jones: Excellent. Can't be as expensive as Australia. I'm paying $440 Australian dollars a month for mine here at the lab.

Chris Gammell: Yeah, that's a bit much. I pay like $40 or something. Oh, God.

Chris Gammell: Yeah, that's not bad at all.

Dave Jones: Go away.

Chris Gammell: So Gary was helping me test this last night too. And so Gary was very nice to volunteer and help us out with call screening, stuff like that. So people who call in will talk to Gary and kind of give their particulars to him. But Gary actually works on ROS, Robot Operating System. Oh, awesome. Yeah. So Gary, why don't you tell us about that a little bit, real quick?

Chris Gammell: Okay. So I work for an Argentinian company, actually, that is called Ecumen. And we do software development for robots. We use ROS as our main tool. And yeah, we have a lot of clients in the US and Canada and stuff. And we do basically software development for their robots using ROS.

Chris Gammell: And so what is the base language of that too? That's the thing. So I've heard about it. I know it's on the PR2, which you have a picture. Your Twitter picture is of you, even though you're not working on that directly, you said. But is it all Java or is it all C? What is ROS actually? What's the base language and how's it all put together?

Chris Gammell: Okay. Mostly it's in C++. That's the core of it, at least. But you have libraries, so you can program it in Python, in C++, in Lisp. There are some work on the Java side also.

Chris Gammell: Oh, cool. Okay.

Dave Jones: There you go. Software. Stuff we have no idea about.

Chris Gammell: And robots.

Dave Jones: And robots.

Chris Gammell: I also do electronics. So I'm actually an electronic engineer. So I do basically robots. I do small, cheap robots for teaching children electronics, programming, and robotics. So, yeah, that's...

Dave Jones: Awesome. I'd love to do something like that, if I had the time. Yeah. We've obviously got a big delay here. I wonder where the delay is coming from. If people don't know, we're using callinstudio.com, which is, that's what we're doing this whole thing with. And, well, some of it, we're still using Mumble. So we're using a mixture of Mumble. We're recording locally as well. And we're using this callinstudio thing to handle all the callins. And we've got a toll-free number and all sorts of things. So there's obviously some sort of delay somewhere. It's... I'm not sure what.

Chris Gammell: Yeah, I think it might be at the actual switch level. So I'm not sure. I mean, yeah, we're kind of traversing a bunch of undersea cables. Obviously, we're all over the world here. But it's interesting because we're going to be having... I mean, people can call in from actual telephones as well, which is kind of different. So...

Dave Jones: I'll tell you what. If somebody calls in from a phone booth, an actual phone booth, you get priority. You get straight on.

Chris Gammell: Yeah.

Dave Jones: I want someone to call in from a phone booth. Anyway. So should we give it a go? Should we try it? I think we're ready.

Chris Gammell: Is anyone going to call in? I think... Yeah. I think the main piece here is whether or not people are going to actually call in. I kind of feel like it's going to be like... Do you ever hear the public radio funding drives? You know what I'm talking about? Yeah, right. Yeah. Like, all right. Well, we're just waiting for another call here. Call it. Yeah, exactly. Call on in today.

Dave Jones: Hey, we've got one. We've got one.

Chris Gammell: Operators are standing by. So... Oh, and we can see their number, can we? Yes, but we will never read... We'll never read the numbers.

Speaker ?: Oh, no.

Chris Gammell: We'll never read your number out. The other interesting thing about this too... So I was trying to figure out what this is all kind of using. I think this is using Twilio on the back end, which is like a... It's basically like a AWS or, you know, like web services type thing, but just specifically for voice. So it's all cloud-based type of thing. Um... So... Right. Yeah, it'll be interesting to see how it actually works out. But the idea is that... So Gary will be able to... You know, people will call in. They'll hear a welcome message. They'll talk to Gary. Give him some of their particulars, you know. And then we'll be able to click through and actually accept callers and stuff like that. So...

Dave Jones: All right. Well, let's see. I guess... So thanks, Gary, for... Oh, yeah. That as well, yes. Thanks, Gary, for doing this. So Gary won't... So Gary will, I guess, leave us now, will he?

Chris Gammell: No, well, he'll be doing that stuff, so...

Dave Jones: He'll be doing it in the background. Right, exactly. But yeah, we won't... Yeah, we won't... We won't hear him. He'll be beavering away in the background talking to our callers. Right, exactly.

Chris Gammell: So it looks like we do have our first caller. Do we... Are we ready? Are we ready to talk...

Dave Jones: And we have details.

Chris Gammell: Our first caller, yeah. This is weird.

Dave Jones: It's Richard. Yep. First cab off the rank from Vancouver. Oh, not a yank. Yes.

Chris Gammell: Hi, Chris. Hi, Dave. Richard, welcome to the Airbow. How are you doing?

Dave Jones: I'm good. How are you doing? Hey, Rich. Yeah. Or is it Dick? It's Richard. Am I allowed to call you Dick? No. No, it's Richard, is it? Okay. Right. No worries. Some people like to be called Dick.

Chris Gammell: Sure, Dave, sure. Okay. Yeah, cool. So I'm working on that. Yeah, you're right. Yeah, I'm working on a battery-powered device that's going to be used in pretty close proximity to 10 to 50 kilovolts, kind of the voltages you'd see in a substation. And I'm pretty concerned with protecting the circuit from induced voltages, especially the sensors input from the sensors.

Dave Jones: Oh, okay. Well, there's... Well, if you're talking about inductive coupling, you want to keep your loop areas small. Yeah, okay. That's one... But you're only talking like 50 hertz, aren't you? Like substation 50 hertz, kind of...

Chris Gammell: Yeah, 50, 60 hertz-ish.

Dave Jones: Whatever, 60 over there, right?

Chris Gammell: Yeah.

Dave Jones: Right. So you're probably more worried about capacitive coupling, perhaps.

Chris Gammell: I'm kind of concerned with not necessarily the... Like keeping the loop area small more. I know that there's going to be induced voltages and mitigating the risks, basically. Kind of like... I would expect to be able to use some diodes to the rails or something, but I'm not sure if that will or won't. Honestly, I haven't tried. But assuming that it does get coupled...

Chris Gammell: Could you tell us a little bit more about the application, too? Like, so like...

Dave Jones: Yeah, I was going to say, is it digital? Is it low-level analog? Because that makes a huge difference.

Chris Gammell: It's low-level analog sensor, but going to purely digital from then on. So going through an ADC and so on.

Dave Jones: Well, you obviously want your ADC as close as humanly possible to your sensor. So get it straight into digital as soon as you can. I mean, that's the obvious, you know.

Chris Gammell: Well, so the whole thing is battery-powered, though. So is it... The whole thing is floating... Is it possible the whole thing will be floating on top of that 10.5 kilovolts? Is it like... Are you worried about just like the transients and the fast drops?

Chris Gammell: Yeah, exactly.

Dave Jones: Oh, so you're saying it's actually connected to it, is it?

Chris Gammell: Oh, no, it's not connected at all. It's a completely separate battery-powered device. Okay, right.

Dave Jones: So it's completely isolated. Okay. Yeah.

Chris Gammell: Just because the Vultures will be swinging so vast, I'm pretty concerned with just nuking all the poor little digital electronics there.

Dave Jones: Well, you're the man here, Chris, because you come from ABB. You're a power man, see, whereas I'm not a power man. I'm a low-level, low-signal guy. Anything over 12 volts DC scares the shit out of me. Yeah, that's pretty much where I'm at as well. So I'm kind of like, okay. Maybe 24, you know.

Chris Gammell: Yeah, that's a healthy fear, guys. I mean, like, really, that's really healthy. It's kept me alive so far. I would say that the main thing, though, is, you know, the swing shouldn't be too big of a... As far as I know, the swing shouldn't be too big of a deal unless you have... First off, if one part of the circuit is ground referenced, you know, outside of the rest of the circuit, because then that will create these huge voltage differentials. And then that whole huge voltage differential, of course, the current's going to find its way through somehow. But I'm not sure if the whole thing's floating.

Dave Jones: If the whole thing's floating, I think it's all about if there's any transients, because I think normal state should be fine. I mean, if you're just in a substation and you're talking, you know, 50, 60 hertz, unless there's, you know, some sort of massive transients that create massive EM fields, then, you know, I don't think there's going to be a problem. Okay, then. Something would be going horribly wrong. You probably wouldn't want to be standing there, you know, pressing buttons on your product. So, yeah, I don't... I wonder. We really need somebody who designs stuff into substations like that. Like, I've been into a substation before looking at some electronic, you know, looking at stuff in there. It didn't seem to be any particular... anything special, though. It was just using regular, off-the-shelf industrial PCs and, you know. Like, off-the-shelf, you know, touch panels and stuff. You know, it didn't seem to be a big deal. So, I think you're going to be fine.

Chris Gammell: Okay, that's great to hear.

Dave Jones: Of course, once again, Chris, we're talking out our ass, aren't we?

Chris Gammell: Yeah, I don't know. How do we do the... How do we... Do we have to do, like, a legal statement now? Like, usually people are... It's so one-sided, you know? It's like, oh, well... Yeah. Hope no one follows this.

Dave Jones: So, I've got to ask, what is this thing actually measuring? What are the sensors doing? I'm just curious.

Chris Gammell: It is actually measuring the electric field around the cables or whatever devices nearby. Oh. So, it is... Oh, right. It's not inductively coupled. I'd be... It's more or less an electrometer.

Dave Jones: Yep, yep.

Chris Gammell: Okay.

Chris Gammell: So, you're off...

Dave Jones: In that case, yeah, I'd be more concerned about actually protecting your analog inputs than the rest of, you know, your power rails and your digital stuff and things like that. I wouldn't be the least bit concerned about that. Unless you're operating down at, you know, the bleeding edge, you know, 0.8-volt rails, you know. No, nothing like that. Ridiculous. Nah, 5 volts. Be a man. 5 volts all the way.

Chris Gammell: Yeah. Okay. I would say the best advice here is probably design really cheap and then, you know, just keep replacing them as the pop, right?

Dave Jones: Oh, come on. That's a quality engineering. That's pretty good. I like that. We can't give dodgy advice. Yeah. Yeah. Right. Modular. Modular. Replaceable. Yeah, exactly. Yeah, yeah. Well, in that case, like... Come on. In something like this, I would... Yeah, exactly. I would actually consider a modular front end because if the front end gets... If there's any potential for the front end to get fried, you don't want to be board level repairing your front end board because that's a pain in the butt. If you can swap out, like, the input ADC module, you know, I...

Speaker ?: Yeah, that's interesting.

Chris Gammell: Richard, do you do any opto-isolation or anything like that as well? I mean, obviously, getting 10,000 volts of opto-isolation as well is kind of tough, but...

Chris Gammell: Yeah, no isolation at all. It's meant to be a pretty small device. And, you know, so far, tests have been okay, but I'm... It's too early on to have done significant testing around those voltages. It's kind of like closest I've got to the testing is, like, you know, a couple of meters away from it. And not really, you know, willing to put my poor little prototype through the ropes just yet.

Dave Jones: I can just picture you, like, having your device on the end of, like, a 10-foot-long wooden pole and you're, like, nervously putting it closer and closer to the transformer, you know?

Chris Gammell: Yeah. Wearing, like, a full, like, isolation suit. Right.

Speaker ?: Okay.

Dave Jones: Oh, that's great. What do we... Come on, we've got to say more about this, Chris. We have to have more advice.

Chris Gammell: I have no more... I mean, like, so I'm a little... I was actually surprised that we could say so much without being up close and personal. Like, I was worried that we're going to have to be like, oh, well, Richard, if you could just send us a scope plot, we'll just wait for that to come through. And then, you know, I do wonder about when that's going to happen. Yeah.

Dave Jones: Yeah, should we do that? Should we have some sort of facility for people to... Well, I guess they can. They can just email us a, you know, an image or something if you'd like. Yeah, feedback at theampire.com always works. It's a lot of time. But that takes... It's all the IP and... It doesn't make for good live radio. Yeah. So how many of these are you making? Is it just like a one-off?

Chris Gammell: Small volume initially, like, probably the next, like, 10 to 25-ish units and see where it goes from there. Like, low volume initially. Yeah. Right now, I've been hand-assembling everything, you know, reflow and so on. But hopefully over time it increases. Right, yeah. But it's definitely, like, not urgent to get to, like, market or anything like that, at least mass market of any form.

Dave Jones: The benefit and also disadvantage of low volume... Hey, I'll start with the benefit of low volume production, like, kind of like testing stuff, you know, like this, like, you know, in the dozens, is that you often don't, you know, you don't have a significant number in order to turn up your more obscure faults, in order to show up more obscure failure modes and things like that. So you can often get away with it. Like, you know, you make 10 of something and they might work fine for years and you think, this is rock stable. It's because you don't have a high enough sample group, you know, for things to go wrong often. Right? Whereas, which can be a good thing, right? Because, hey, you can, okay, job done. Wipe your hands off it. Right? And yeah, right. Excellent. Big tick on my performance review, you know, at the end of the year. Right? That's okay. But if you do high volume stuff, then of course you get all these significant numbers. But unfortunately, you only get, you only find out until you've actually produced, you know, 10,000 of them. And then you find, oh, geez, I'm getting, you know, a 2% failure rate. That's not acceptable. Yeah, exactly. Recall time. And that's a world of hurt. But the good thing is it makes a more robust product in the long run. So if your goal is ultimately, if you've got a 10-year lifespan product, you want, you know, you don't want to be making 10, you know, you really want to be making 10,000 and flesh out all the problems. But yeah, but that can be embarrassing, you know, if there's a large failure rate. So those annoying ones is where you're not sure why, you know, I can imagine this being one of the scenarios where, you know, you get a report back from somebody in the field, you've made 20 of them. Oh, it failed. And you get it back and you go, well, okay. Yep. Why? You know, without actually having sensors and everything in place in your actual test, the real test scenario, you can't tell. Like, you know, you can't tell what caused the failure often. And that can be a real, you can be chasing your tail forever. It's not like something like this when you're right next to a substation transformer. You can't just walk out for multimeter and start probing around.

Chris Gammell: Initial test units will actually have a bit of sort of data logging in them to send us back some data that's been gathered on how they're being used and what kind of things we're picking up. Smart man. Hopefully, if they don't explode and catch fire, you know, send us any faults.

Chris Gammell: I was going to say, what if the data that's sent back is just, ow, ow, stop it. Why is it such high voltage? What do you mean?

Dave Jones: There's actually, I'm judging the Hackaday Prize at the moment. Sorry for the little segue. I'm judging the Hackaday Prize at the moment. And there's this project in there where it actually generates a human voice and actually sends it back via RF so you can listen on an actual radio so you can actually hear it. It's talking, you know, sensor at 102% alert. You know, like, it's great.

Chris Gammell: Yeah, that's awesome.

Dave Jones: I think we should incorporate voice into our products more. You know, you get these cool voice chips now, you know.

Chris Gammell: Well, we're passed back to the future today. Isn't that what we're supposed to be at? Everything speaks to us and, you know, you've got computer that you just say, computer.

Dave Jones: That's right.

Chris Gammell: That's right.

Dave Jones: And everything speaks like Max Hedrum, you know.

Chris Gammell: Yeah.

Dave Jones: From the Cafe 80s. Sorry.

Chris Gammell: If you don't know the show. Yeah, I remember, I think I've mentioned on the show before, but I was always really keen on the idea of putting that into test equipment. I tried pitching that when I was at Keithley. Right. Because you think about when your head's down on a board, so you're, like, not even working high voltage. Oh, yeah.

Chris Gammell: Trying to probe around. Exactly. Yeah, yeah.

Chris Gammell: Like, what the reading is. And, you know, like, if you don't have that, you're kind of...

Chris Gammell: Actually, I've seen some multimeters that do have that. They'll, like, read out aloud the, like, voltage reading. They're not on any scopes. Yes. I've seen that in some multimeters somewhere.

Dave Jones: Oh, okay. There's a couple of them, yep.

Chris Gammell: Well, then, maybe I'm wrong. Maybe not.

Dave Jones: It's a very niche thing, you know. Yeah. You've got to want to have that.

Chris Gammell: Maybe someone was pitching to the show and they were, like, done.

Chris Gammell: Yeah, right. That's how it's done, yeah. So, we're like Shark Tank in reverse. Right.

Dave Jones: Well, have we answered your question, Richard?

Chris Gammell: Yeah, yeah. Kind of. Sort of. Not really. You definitely relieved some of my concerns and, yeah.

Dave Jones: Yeah, no, I think she'll be right as my official. As my official Aussie response. She'll be right, mate. No worries.

Chris Gammell: Sounds great.

Dave Jones: So, you're actually Canadian?

Chris Gammell: I'm Australian, but I've been living in Canada for a long time now.

Dave Jones: I thought so. I thought that was an Australian accent, but I wasn't sure because of the poor voice quality of this connection, you know. I thought, yeah, from, like, your first time you spoke, I thought, Australian? What?

Chris Gammell: Yeah, Adelaide.

Dave Jones: Ah, sleepy Adelaide.

Chris Gammell: Yeah.

Chris Gammell: How's that, Chris?

Dave Jones: You're surrounded by her.

Chris Gammell: That's okay, man. I've gotten used to it, so.

Dave Jones: Alrighty. Thanks for the call. All right, well, yeah. Thanks, Richard. You're number one. You could be the last.

Chris Gammell: Oh, God. Well, hopefully more call up and get through, and I look forward to listening.

Chris Gammell: If not, we'll do our usual BS that we do every week, so we'll be fine.

Dave Jones: Yep, we'll be fine. Awesome.

Chris Gammell: Thanks for calling in.

Dave Jones: Thanks, Richard, for being the first guinea pig.

Chris Gammell: Awesome.

Dave Jones: See you, mate. Bye. All right, well, that was good. Man. That worked. And the delay, there was hardly any delay there, so I think Gary's, it could be something to do with his connection.

Chris Gammell: Oh, I was getting a little bit of delay, but yeah. Oh, really?

Dave Jones: Okay. It seemed pretty good to me. Insider baseball. That's okay.

Chris Gammell: It was quite small. Yeah. So, you still haven't ever listened to Car Talk, have you? Right?

Dave Jones: When you first told me about it, I went and had a listen, and it's just so not what I'm interested in. See, that's the thing.

Chris Gammell: Those guys were troubleshooters, though. That's what's amazing about it. And it's like pattern recognition stuff like that.

Dave Jones: Yeah, you've got to listen to the whole show. Yeah, you've got to listen to probably a couple of shows to get the feel for it, and I wasn't, you know, I'm not a big podcast listener. Right, exactly.

Chris Gammell: So, yeah, so that's definitely one thing.

Dave Jones: Sean on Twitter is asking, are we doing a live show?

Chris Gammell: Yes. Well, sort of. Not really.

Dave Jones: Well, we are.

Chris Gammell: Yeah.

Dave Jones: I'll call in and find out.

Chris Gammell: Yeah, so the thing that I was going to mention, though, is that on Car Talk, those guys do something called Stump the Chumps, and they would call back in, like, later, and they would see if they were right.

Dave Jones: Ah, yeah, right.

Chris Gammell: I don't know if we were able to do that with Richard, though, because he might be 10,000

Dave Jones: Six months down the track, we'll have to call back and say, yeah, every 10, 10% of them just blew up, you know?

Chris Gammell: Yeah. Well, I mean, if he blew up, though, that's the problem. That's the real issue.

Dave Jones: I mean, see, here's the thing. Like, you want to give advice for something like this, but my first gut reaction when he told us, oh, you know, he's putting a sensor next to it, and it's floating next to a power transformer. It's like, you'll be fine, you know? No wuckers. Yeah. So, but that's not really, you know, that's not really solid engineering advice, is it? It's not.

Chris Gammell: I mean, the things I think about are, like, capacitance between the, you know, power and ground type of things. If you could have those transients just even between, you know, small transients between the board, if there's a big enough drop, that kind of thing. Like, I don't know if that's, if that would actually happen, but, you know, floating electronics is awesome. It's just super hard to, actually, the tough part is measuring, because almost every scope and every bench DMM, you have to get everything handheld, because any kind of ground reference piece of test equipment.

Dave Jones: Yeah, mains earth reference, yeah. It's bleh, bleh, bleh, bleh.

Chris Gammell: You will ruin.

Dave Jones: Yeah, I've done a video on that, how not to blow up your oscilloscope, and people, and I don't blame them. People still have a hard time understanding the concept of what's going on there.

Chris Gammell: Yeah, right.

Dave Jones: You know, and yeah, you've got to think about the whole system, and where current is flowing, and, you know, references, and how, you know, ground isn't ground, and, you know, well, ground can be anything you want it to be, and, you know, like, we do have a couple of more callers. We do.

Chris Gammell: So, let's welcome Arsenio.

Dave Jones: Arsenio. How you doing? Hello, Arsenio.

Chris Gammell: Hey, Dave.

Dave Jones: G'day.

Chris Gammell: All right, so I'm having a problem with these, uh, yeah, g'day. All right, so I've had a problem with these I2C lines.

Chris Gammell: Yeah.

Chris Gammell: So, they keep pulling low for some reason, and I don't know why.

Chris Gammell: Oh. What are they connected to?

Chris Gammell: Uh, they're connected to Arduino and an IMU. They're currently, uh, connected, uh, in my rocket stabilization system prototype.

Chris Gammell: Whoa. Okay. Um, we might need a link to that eventually, because that sounds awesome.

Dave Jones: Uh, so you're saying that...

Chris Gammell: It was all hacking a while back. Okay.

Dave Jones: All right, so you're saying that they're pulling low. They're pulling low even when they're not supposed to be communicating. Is that what you're saying?

Chris Gammell: Yeah, that's correct.

Dave Jones: So... Okay, is it...

Chris Gammell: I've checked on my crappy logic analyzer, and it shows that there's a packet ping, and then it just doesn't respond. I'm like, well, that's weird. So, I disconnect the logic analyzer. I run it again. It queries. About 80% of the time, it responds. But the other 20% of the time, it doesn't respond.

Dave Jones: Huh. Oh. Oh, I squared C is... I like it, but it can be a pain in the ass to troubleshoot sometimes. Well, what's your... Is your IMU...

Chris Gammell: I'm debating whether I should go with I squared C or SPI. Or is it SPI? I don't know.

Dave Jones: Well, SPI, SPI. Yeah, either side, I think. Some people like to call it. Yep. I... Yeah. I mean, SPI is more robust. You know, I squared C, you've got the capacitance of the line. If you don't choose your correct pull-up resistor, you can get interference, blah, blah, blah. You know, like, it's... It isn't as robust as a good, solid totem pole output SPI driver. You know, boom. There it is. You know, none of this pull-up rubbish.

Chris Gammell: Well, that's a new one. I've never heard you described it as a totem pole.

Dave Jones: Well, SPI is a... You know, well, all digital outputs are totem pole outputs. So, that means they have a driver and two output transistors. One pulls it high and one pulls it low, whereas I squared C is the opposite. I squared C is an open collector output, essentially. And to pull the line high, you're relying on that pull-up resistor. So, your pull-up edge, if you actually have a look on your scope, probe the line, and zoom right in on the edge, and you'll see that it'll pull down very quickly. It'll pull down, like, almost instantaneously, right? Because there's a transistor there to pull it down.

Chris Gammell: I don't have a scope.

Dave Jones: Oh!

Chris Gammell: Ah, we found the first step. Oh! There you go. That is a big step. I mean, it really is. It is a big step.

Dave Jones: That's the other thing. A lot of problems... You're a budget here. A lot of I squared C problems come down to you. You've chosen the wrong pull-up resistor value. What value are you using?

Chris Gammell: Right now, I'm currently switching between a 10K and a 5K. Neither work too well. Right. Okay. I was like, let's try 5K. Didn't work.

Dave Jones: I'd probably lower that. What should I be using? Yeah. Like, I believe, like, the industry standard value is 2K. That's, like, you know, recommended. But you can even lower it down to a K. You know, lower it down to 1K and see if it makes a difference.

Chris Gammell: Do you understand the relationship, too, with the, you know, higher or lower resistance and how that ends up affecting the whole circuit overall?

Chris Gammell: To the most extent, yes. But I'm still pretty fuzzy on it because I'm doing this with very, very little actual knowledge. Okay. Right. It's more or less flitzing about with it to, like, get it to work.

Chris Gammell: Yeah. That's understandable. I mean, so the main thing is that, like, so the idea of a pull-up or a pull-down is basically, like, using the rail or using ground to inject current. So that if you think about it as, so you said you're using Arduino. Like Dave said, it's an open collector. So what's happening there is really all it's doing is usually tugging down. And you have a pull-up resistor there so that when the Arduino is no longer pulling down, instead of the voltage drifting higher and higher just because of, you know, just leakage currents or anything like that, what you do is basically you have this pull-up resistor contributing current from the power rail, but you don't want to have it actually shorted to the power rail. You want to have, you know, some kind of controlled amount of current being injected into that line, and that helps to bring that voltage up quicker. So a lower resistance for the pull-up resistor ultimately ends up injecting more current.

Dave Jones: Which is better. So a lower value pull-up resistor, the better it is, and the faster it's going to pull it up.

Chris Gammell: Right.

Dave Jones: So the less timing issues you might have or something like that.

Chris Gammell: And the trade-off there, too, is that the reason people use higher values in the first place is power, right? Because if you think about it, if you're using, like, a super low-power device, and you could use a 10-ohm resistor if you wanted to, it would be, you know, but then you end up dumping a ton of current into there through that pull-up resistor, and, you know, either you're going to have to, you're going to burn a bunch of power directly, or you're just going to burn your battery out because you're going to be using so much current in order to simply raise a lineup. So usually it's a higher value, like, you know, 2K, 5K, 10K type of thing. So, yeah.

Speaker ?: Yeah.

Dave Jones: But something like the SPI bus, you really should get a scope. It'll really tell you what's going on there.

Chris Gammell: Yeah. I'll have to be at least.

Dave Jones: Yeah.

Chris Gammell: That is a big one.

Chris Gammell: All right. So another thing is I had a strange bug that I fixed by changing the baud rate from 3200 to 115200. What was happening was it either randomly froze up with the FIFO bus overflowing, or it did what it called a Harlem Shake glitch, where even though the sensor was absolutely still, and the silicon had come up to temperature, it just spewed random values like it was turning itself inside out and flipping on like a doomed progress ship.

Chris Gammell: Very illustrative. Yeah. You have a gift with words, my friend. Thank you. Yeah. You know, I am allergic to phrases like, what did you call it? Like, it had a bunch of noise, or it, what was the, how did you say it at first? No, it just put itself inside out. Yeah.

Chris Gammell: Wait, the first one? Yeah, the first one. Oh, that was a FIFO overflow for some reason. Yeah. But increasing the baud rate fixed it. I don't know why.

Dave Jones: Hmm.

Chris Gammell: That one's kind of tough to troubleshoot with.

Dave Jones: Yeah, that's, that's like a software, yeah, that's a library issue, how your comms library is written and how it, how it clears the FIFO, you know, how it actually clears the buffer and all that sort of jazz.

Chris Gammell: I think another thing to point out is if you're not setting the baud rate the same on either side, you start to basically, you know, you're not, so basically with a serial link like that, it's expecting data at a certain rate, and if it doesn't get that, then it's all just going to be garbled packets in the first place. So, that would be a big part of it.

Dave Jones: So, do you have the opportunity to change to a spy bus for this sensor?

Chris Gammell: Yes, I do. It's currently all on just breadboards and proto-boards right now. I'm moving towards, slowly I report actual DVDs.

Dave Jones: Yeah, that, that's the other thing. When you've got things, when you've got breadboards and wires hanging everywhere, you've got capacitance and you've got a big antenna, basically. And when you've got an I-squared C, that is not ideal for an I-squared C line. That's why you want to have a lower value. Whereas an SPI line, as I mentioned, is a direct drive totem pole output. So, you've got no issues with pull-up, pull-down. You don't need them. You don't need a pull-up resistor on a SPI line because it's just driving directly. And it's going to be more rock solid.

Chris Gammell: That might be a bit of a red herring, I think, in general, though. It could be. There's tons of potential issues. I mean, there's addressing stuff like that with I-squared C. There's tons of issues like that. Yeah. There's a lot of good tutorials out online. Actually, my friend Dan over at Reingold Heavy does a pretty good I-squared C SPI set of tutorials. We can put a link to that in the show notes. Yeah, I've had a reading stuff recently. Okay.

Dave Jones: Well, if I build up something and an I-squared C line doesn't work, right, and I'm sitting there scratching my head, why does it work? The first thing I'm going to do is actually get the scope out and just probe it to make sure that there's nothing funny with the pull-up or anything like that. That the edges are nice and clean and everything's talking. And then, after you've checked that, then you go into the chip addressing. Is the chip being addressed? Is it enabled? All that sort of jazz. And then you get into the protocol. Well, is this a proven I-squared C library? Have I used it before? Or did I download it off the internet? And I've got no idea if it works.

Chris Gammell: Yeah, and if you use like a CELIA type of thing, or I mean, there's a lot of them out there, but you can actually do the translation in the logic analyzer a lot of times too. And that is just magic.

Chris Gammell: Yeah, I've got the $10 clone one.

Dave Jones: Okay. Oh, okay. Right.

Chris Gammell: I feel kind of guilty about getting it right now. Mark and Joe are going down the line. I'll go and get an actual one.

Dave Jones: Well, you know, like, yeah, as long as you go legit later, then.

Chris Gammell: I'm intending to get a real one. Yeah. Right.

Dave Jones: Exactly. Excellent. Then that's, then that's, I think that's fine.

Chris Gammell: Future customers. Yep.

Dave Jones: Yep. Cool.

Chris Gammell: Arsenio, thanks so much for calling in. We really appreciate it. We got another call, but good luck with those issues. And, you know, let us know, let us know how stuff goes.

Chris Gammell: I'll link you guys to my project on Twitter. All right. We'll keep you posted. Thanks for having me on.

Dave Jones: Thanks. Thanks, mate. See ya. This is working, Chris. It's kind of working.

Chris Gammell: Yeah.

Dave Jones: That, the, uh, Arsenio's voice quality there was a bit dodgy though. See, I thought it was okay. I don't know. Yeah, I guess. Oh, okay. No, it sounded very.

Chris Gammell: We love feedback. People can leave stuff in the comments. I'm going to clean some of this stuff up in post, I'm sure. But, uh, but, uh, yeah. Let's welcome our next caller. We may have a, uh, uh, a voice that we recognize here. Hello. You're, you're on the amp hour.

Chris Gammell: Hey, Chris. Hey, Dave. This is Brandon Harris.

Chris Gammell: Our former guest, uh, from the electric amp. How you doing, man?

Chris Gammell: Good. Good. How are you guys?

Dave Jones: Good. Well, we're surviving. People are calling in, which is good. Yeah. Our first call-in show. Uh, it would be really embarrassing. We're just sitting here going, oh, I don't know.

Chris Gammell: What people don't know is we beg Brandon to call in or something, right? Right. Yeah, right. Yeah.

Chris Gammell: Please call us again. This strays a little bit outside of strictly hardware and more into, uh, sort of culture and thought leadership. Um, you know, one of the big things that's happened with, in software, um, with security things is, uh, the notion of responsible disclosure. And I was kind of wondering what, what you guys think of that in terms of how it translates to hardware engineers, because we're increasingly getting to the point where, you know, hardware, uh, is, is a significant part of the security story, you know, whether it's even in some cases, very simple thing with a garage door opener, um, you know, up to very complex things and, and IOT enabled things. And, and whether, you know, as hardware engineers, we have a responsibility to disclose that to a manufacturer rather than choosing to publish, um, that on the day and, um, you know, sort of pull their pants down in public. Right.

Dave Jones: Right. Okay. So you're talking about, um, like there's, there's sort of, yeah. Okay. Isn't there like a sort of unwritten rule in the security hacking world that if you find a hack to a product, let's say you figured out how to hack some sort of, you know, the new credit cards or something or a car or something, you know, would you, yeah, like, you know, you're supposed to contact the manufacturer first and give them what, two weeks to respond or something. And if they don't, then you go public. Is that what you're talking about?

Chris Gammell: Yeah. So, you know, for example, uh, I, I not so long ago spent some time reverse engineering a product and found, um, that the hardware, the wireless protocol they use is wildly insecure. Right. Um, and there are people doing things like, um, putting front door locks on this, um, and using it for garage door openers and things like that. And, you know, do I have a, one of one disclosing that to that manufacturer? And, you know, do I have an obligation to not publish it until I give them a chance to respond? And it's obviously different with hardware and software because, you know, in software, if I find a bug on a server, they can go fix it, but they may not be able to come out and fix it.

Dave Jones: Right. Hardware is tough. Yeah. Yeah. Yeah. Well, so do you, yeah, do you, see the, the cynic in me, right? Me cynical. Who would have thought? Um, the cynic in me would always think that if I go tell the manufacturer, they're just going to sweep it under the carpet. They're not going to say a damn thing. It's almost as if I would be inclined to like, you would have to threaten them. It's like, if you don't, I'm, I'm reporting to this, I'm giving you an opportunity to respond. But if you don't, I'm going public and I've got 280,000 YouTube subscribers and you're going to be toast, dude. You know, like, yeah, as somebody who's worked in large companies like this and, and seeing things being swept under the carpet, I mean, it's just a company's natural response, especially as you said, because they can't easily fix it. Right. They can fix software. Oh yeah. Patch. Right. Done. Send it out. Right. Fixed. Not a problem, but hardware. How do you, you know? Yeah. It's got to be a case by case basis. Doesn't it?

Chris Gammell: Disclosing it may, you know, may hurt the public in general. Right. You know, that it may, it may help. It could hurt the public and the company.

Dave Jones: Yeah.

Chris Gammell: And, and so, you know, it's, it's, it's kind of a moral dilemma, right? That software is fairly well defined. And I thought you guys might have some, some thoughts to share with the community.

Dave Jones: Well, I am, I am not a sweep under the carpet guy. I am a, I'm going to stand in the camp of, well, you know, it's there. Don't sweep it under. It's going to come out eventually. You might as well do it and fix it. I don't care how much pain it's going to cause.

Chris Gammell: Well, I think there's an interesting piece there. I mean, like, so this has existed for a long time, right? I mean this, but it usually exists more, it's more front loaded. It feels like because like, this is what errata is. Like if you have a new piece of silicon and you're willing to use a new piece of silicon that is not, you know, battle tested by hundreds of engineers, then yeah, you're going to be probably, you might be potentially discovering silicon errata, right? And that's the idea. And that's what errata notices are for. I don't know. At that point, do you at least turn to your coworker and you're like, hey, I wouldn't use that chip. You know, that's kind of risky.

Dave Jones: Yeah, but a chip, it's not a product, is it? That's the difference. I think there's quite a potential bit of difference there between reporting a bug in a chip and reporting a product as such that every, like that, you know, consumers, Joe Average might be using, you know?

Chris Gammell: Yeah.

Dave Jones: And that hackers can take advantage of, you know, if there's an errata in a chip, it's like, well, you know, like that's like companies should have tested their own products who built with that chip, should have tested their own products to make sure it works, you know? And it's probably not a problem for them because they've tested it, right? Yeah. So it's less of an impact.

Chris Gammell: To play devil's advocate, you know, at Electric Gamp, we are, you know, we start with security at the hardware level and we can, you know, we can test a chip to verify that it is secure in the way that they say it is. But, you know, actually, intensive, you know, security validation and, you know, oh, there's a buffer overflow in the bootloader, right, that can be executed, right? I mean, you know, this is something we also dealt with at Apple where, you know, everybody and their mother is trying to jailbreak an iPhone. Oh, right, yeah. And, you know, whether it's hardware or software, right, you know, you're still beholden to these, you know, insane cases. And, you know, in some cases, the manufacturer may be able to mitigate it. In some cases, they can't. But, you know, it's kind of a, to say that, you know, Apple surely can test, you know, all of their software and all of their chips. And yet, iPhones are typically jailbroken, right? So, it's kind of, it trivializes the problem to say you should just test your product.

Dave Jones: Can I put you on the spot here? Because you called in. Got him. We got him, folks. Yes. And being a manufacturer of a product like this, that, I, sorry, I don't know if the security things in the imp itself. But let's say, okay, you've manufactured your million imps, right? And they're gone out there. Everyone's using them. And someone found a security flaw in it that makes it possible for anyone to hack into any electric, secure electric imp anywhere. And this could potentially be a massive recall for you. As the manufacturer, what would you think? Would you thank them? Oh, thank you. Or would you go, oh, shit, we're going out of business. I'm going to sue this guy, you know, for actually releasing this. As a, I know it's a bit of a hypothetical.

Chris Gammell: It's almost like a libel type thing, right? I mean, it's almost like that.

Dave Jones: Oh, it's almost libel. Yeah. Potentially.

Chris Gammell: Yeah.

Dave Jones: I mean. You can be sued at any point for anything. So, you know, yeah, if you did this to a big company, right? Anyway, sorry, go ahead.

Chris Gammell: I can't speak from an electric imp standpoint. I can speak from mine and what I would expect electric imp to do. You know, I think if I think I agree, right? You have to step up and, you know, put on your big boy pants and own up to your mistakes, right? That that's part of it. You know, we've spent a tremendous number of hours working on it, protecting our system. It's something that, you know, we consider part of our core philosophy. We've hopefully, you know, we also do things preemptively. I did work at Apple. I do. My expectation is that no matter what I do, somebody will find a hole in it. And what you need to have sort of, you know, we talked about like layers of the onion, right? And so in some cases, you can play the cat and mouse game where it's like, okay, well, I can make this hard enough for somebody that, you know, it's not worth their time. And maybe at one of those steps, I blocked them completely, right? So you always have multiple levels of security. And, you know, we would have to look at those things. But I mean, yeah, I mean, absolutely. Like, it keeps me up at night, right? Like, it would be devastating. It would be, you know, I would be extremely personal. You know, that it would feel like a failure.

Dave Jones: But that's a risk, isn't it? Yeah. Of being in the business that involves security, any form of security like that. You can't even, your best intentions, you could test it until the cows come home. But, you know, nothing's 100% testable. You know, somebody might always find a flaw. I mean, that is a risk of being in that business.

Chris Gammell: The phrase we use a lot around here is there's no such thing as a secure system. There's a system that is secure at a point in time. And, you know, everybody thought their servers were secure and that they had all the latest SSL updates. And it turns out they all had a massive heartbleed, you know, bug. And nobody found it. Nobody knew about it.

Chris Gammell: Oh, yeah, I remember that.

Chris Gammell: It was super tested. But, like, I mean, banks and stuff had huge holes in their security. And they've been running that way for months. And nobody had caught it. And, you know, it was great to see the community respond. Like, I mean, that thing got announced. And there were patches day one. And servers were being rolled. I mean, it was great to see that level of response. But, you know, I mean, it's true that, like, you can't design a system that is secure today and secure in the future. Right? You can do the best you can today. And you have to give yourself the tools to improve it in the future.

Chris Gammell: Cool. I've got nothing more to add. Yeah. I mean, unfortunately, I think, Brandon, unfortunately, the security side of things, it's like, I don't know, like, how much is hardware really at the security level as well? I mean, like, I think about it, like I mentioned, Arata, but security, like, you know, there's encryption level. Like, if the silicon's busted for the encryption, you know, part of the chip, it's like, I don't know, man. Like, that's like a recall at that point. That would be totally worthless, right? There's no, is there any fixing that? Like, if it's spitting out bad hashes?

Chris Gammell: I mean, there are certain things that there's no fixing, right? But things like, you know, we're getting more and more tools, like the chip whisperer is a great example. Yeah. That's another one of the things where it's like, yeah, your design is secure today because nobody can look at, you know, nobody can capture the power line traces. But all of a sudden, the chip whisperer comes to market. Everybody and their brother can start doing, you know, power band analysis and all these side channel attacks and your system, your security breaks down. The people who are going to be doing those attacks are going to be hardware engineers, right? The only people that understand that stuff enough, you know, are going to be hardware, you know, and with some software. But that's the kind of direction that I think most hardware engineers are moving now, right? You know, even you, you're writing, signing micro-cricklers and writing software and stuff.

Dave Jones: In the past, it has required hardware knowledge to do that. But now, if you can buy a chip whisperer off the shelf, you can have no hardware knowledge and still potentially, or little hardware knowledge and still potentially, you know, do these sort of attacks.

Chris Gammell: And it's up to the hardware engineers to try and prevent them and fix them, right?

Chris Gammell: Yeah, I mean, I think the best people that are preventing it are trying all this stuff themselves. They're, you know, they're white hat type hacking, that kind of thing, right? I mean, like, they're testing their own products like that. Yeah.

Dave Jones: All right. Well, thanks, Brandon. I think we're going to have to talk more about that. Thanks for talking to you again. You know. Yeah. Well, thanks, guys. Interesting topic. It was a little outside the hardware realm. Cool. That's good, though.

Chris Gammell: Thanks for that, mate. I think that's definitely a good topic. And I'm sure we'll get some good comments in the comment section. And people can always write feedback at theamphour.com.

Chris Gammell: Yeah. Good luck with the rest of the show.

Chris Gammell: Thanks, man.

Dave Jones: Thanks, mate.

Chris Gammell: Bye, guys.

Dave Jones: See ya.

Chris Gammell: That was interesting. Yeah.

Dave Jones: And we've got another caller. Oh, they're coming thick and fast. This might be our last caller.

Chris Gammell: I think it might be. Yeah. So let's pick up with Steve from L.A. And he's wondering, how do youngsters get started without using SMD parts, especially as they get more common? Steve, welcome to the show.

Chris Gammell: Hey, Chris. Hey, Dave. How's it going?

Dave Jones: Hey, Steve. Good, mate. What sort of accent can I hear there? Oh, no. I'm just from Los Angeles. So I'm a Californian. Californian. Yep.

Chris Gammell: Right. So you guys were talking about the demise of the through-hole parts. And I teach a science class for third and fourth graders, and we use through-hole parts, you know, occasionally and whatnot. And what happens if they all go away? What are we going to learn from? Surface mouth?

Dave Jones: They're not going away. Through-hole parts are not going away. They still have their place. It's, you know, people have been saying that for 20 years, but it hasn't happened. And there's no signs of it happening. So, you know, I don't, like, yeah, it's getting less and less, but there's still hundreds of thousands of them out there. So I don't think it's a major concern.

Chris Gammell: Do you think it's, I mean, yeah. And I mean, I guess one thing is that you're always going to be able to find new old stock type of thing, right? Just because of the scale of the industry. So that'll definitely help. And I think that even it's still cost-effective on some levels to do, you know, through-hole resistors, or there's power requirements that require through-hole resistors and stuff like that. I think really it's the more, it's the newer, more advanced chips. But in those cases, you know, if you're getting started, you're not going to be jumping right into doing like, you know, the 5-amp buck boost or something like that. You're going to be starting a little bit smaller. Exactly. And 555, stuff like that. You're going to, you know, they're making tons of money on those parts. They're, you know, a cost of cent to make and they sell it for five. It's like, okay, well, they're still making a healthy margin there if you make it up in volume, right?

Chris Gammell: Yeah, because I started out when I was like nine years old. And, you know, we just make airborne circuits and things are so easy. But now if I have like controllers and stuff that I want to use, if I, I actually go with the team fee and stuff like that, that have the, that they're actually needed. So we can actually still have access to those boards and stuff. So I don't know. It was just a thought. It was so easy to learn when you had, you know, circuits flying out in the air and soldering and whatnot. It's just easier than for youngsters.

Dave Jones: Yeah. Well, you can still do it, but there's, you know, so many, but there's so much more advanced stuff these days, which is only available in SMD, as Chris mentioned. Yeah. You know, and that's just, well, you know, that, that, that's just the way it is, but you know, there's still infinite amount of stuff you can learn with, uh, through whole parts. So, you know, yeah, sure. You can't get an, an accelerometer, for example, in anything, but one of those pain in the ass leadless chip carriers, right? But Hey people, but for, but for a couple of bucks on eBay, you can buy one. Somebody's already sold it onto a dip adapter and plug it straight in. Just plug it straight into your breadboard or, or wire it into it. I mean, you know, so it's always going to be the 10 years, 10 years, all this stuff is still going to be around. Yep. I bet your life on it. Bet your life on it. All right. Yep. Like if you go, can you, can you go to digi key now and search for all through whole parts? You probably can in, in terms of chips, right? Go there and just search how many are available in a dip package, for example, and you'll probably get tens of thousands. Parts IO. No, Parts IO. All right. Mr. Parts IO company. Right. You know, there's tens of thousands and you know, 10 years ago, for example, there were still tens of thousands. Like I, you know, yeah, it's dropped a little bit, but I don't really, you know, no, I can't see it suddenly vanishing and dropping off the cliff in five to 10 years time. Not going to happen. Don't worry. No worries. She'll be right. That's the second time I've said that in today's show.

Chris Gammell: All right. Very good.

Dave Jones: That's my answer for everyone.

Chris Gammell: Cool. Thanks for calling in, Steve. We appreciate it.

Dave Jones: Thanks, Steve. All right, Chris. Bye.

Chris Gammell: All right. Well.

Dave Jones: And that's, we don't really have any more, do we?

Chris Gammell: Uh, no, but that's okay. Uh, this is our first time.

Dave Jones: Anonymous caller, but, uh.

Chris Gammell: No, no, that's, that's Gary. That's our, that's our screener. Oh, that's. Yeah.

Dave Jones: Oh, oh, sorry. Screening room. Right. Yeah. Okay. Sorry. I'm still getting used to this software.

Chris Gammell: So what do you think, Dave? I mean, should we do this again? I mean, yeah.

Dave Jones: Well, it's worked. We had no crazy people call in. Yeah. Well, maybe we did and Gary, and Gary gave him the boot. Yeah. That'd be good. Uh, yes. Yeah.

Chris Gammell: Uh, I, I think these are all actually really good, good questions and everything too. You know, like, uh, easily describable issues, stuff like that. Actually, I just saw Phil Frieden, one of the, uh, uh, Frieden, sorry, the, uh, former guest was making a joke on Twitter about, you know, if, if there's like really hard to describe conditions, then it's not going to work as well. But, uh, you know, if there's, if there's a lot of, it's going to be kind of us talking about basics, like pull-up resistor, stuff like that. I think that that's good though, because it's, it's the stuff that you and I might not talk about on a normal basis. Right. But a lot of people have a question about it. So that actually works out really well.

Dave Jones: Well, I expect most questions from, from beginners, like, you know, like as, um, sorry, who was it? Richard? No. Uh, the, uh, right. With his I squared C pull up. It didn't have a scope. Right. Right. It's like kind of, you know, that's where I expect the majority of questions to, uh, come from. Yeah. Rather than, you know, experienced industry guys and, you know, I've got this, you know, a thousand pin BGA and I'm doing DDR5 SRA and I'm having issues with, you know, it's pin G11. Just fix that one.

Chris Gammell: That, that, that, that's, that's the answer. Yeah.

Dave Jones: Right. Yeah. I don't expect, uh, too many of those, but I like the spontaneous nature of it. You know, I mean, you know, people get our first gut reaction to questions.

Chris Gammell: Well, Dave likes is not having to prepare. That's what he's trying to say.

Dave Jones: I do enjoy not preparing. Well, it changes things. Like if we had a bunch of questions, it, it'd change the whole look, not look and feel of the show, change the whole feel of the show, you know, because we'd, we'd have prepared answers that we'd want to go through. And like, it, it just, yeah, it might be a bit more accurate and, you know, and, uh, and thorough.

Chris Gammell: People are always in all this accuracy and right answers. Yeah, I know, but it's not fun.

Dave Jones: Yeah. Oh, it's, I, yeah, no, I think the spontaneous nature is, is good. People get our gut reaction to things.

Chris Gammell: Yeah. I think, uh, next time we'll try and make it a little bit quicker. The, the response time is probably the big thing. So, but.

Dave Jones: What, as in us answering the question?

Chris Gammell: No, no, no. I mean like the, uh, the delay in the line. That's just because it's all soft. If we switch over to a hardware solution and I do have one person that might be, Charles might've said he could help us with that. So we might be able to do that eventually. Okay.

Dave Jones: Jeez. We've only used calling studio once and you're looking to ditch it. What?

Chris Gammell: Maybe. Uh.

Dave Jones: It seems to work, but yeah, I mean, you know, it's always going to be dependent upon the phone quality at the other end, you know, I mean.

Chris Gammell: Right.

Dave Jones: Or, or what they're using. Are they using a direct landline? Are they using Skype in.

Chris Gammell: Yeah.

Dave Jones: You know, Timbuktu.

Chris Gammell: Yeah, man.

Dave Jones: So. No. But anyway, it worked.

Chris Gammell: Yeah.

Dave Jones: I don't think we'll be doing this every show.

Chris Gammell: I'll tell you that. Uh. Oh, no, no, no. This will be a, this is a sometimes food, you know.

Dave Jones: Right. Junk food. Right.

Chris Gammell: Exactly. Uh. Uh. So I think next week we'll be back just doing our normal thing. I will be, I think we'll be, well, I'll be, uh, heading to, I'm going out to the conference next weekend. So.

Dave Jones: Which conference? Oh, the, um, yes, the Hackaday conference where they are announcing the winners to Hackaday. That's right.

Chris Gammell: Yeah. And you are off to finish your judging right now, you said, right?

Dave Jones: I'm finished. Yeah. Yeah. And Sophie's, um, pestering me. When are you going to finish the results? Right after this show. So it's your fault, Chris. It's my fault. Yeah. She'll yell at me. Yeah. It's fine. Yeah.

Chris Gammell: Uh. That's good. Yeah. No, that's exciting. And I, yeah, I think, uh, you know, there'll be a lot of familiar faces out there. So I'm looking forward to that.

Dave Jones: Yeah. Can I mention something about the Hackaday judging?

Chris Gammell: Sure.

Dave Jones: Um, it's, you know, I've, I've got to give a score. I've got it. Like I've got, as the other judges have, how many judges are there? There's like eight.

Chris Gammell: I think there's more this year, like 12, maybe more 12.

Dave Jones: Okay. Yeah. There's a lot of judges. Anyway, we've got a rank. We've all gotten the, well, I believe, correct me if I'm wrong, Chris, we've all gotten the top 10, right? So it's top 10 projects. Yeah. I don't know. Pretty sure it is. Anyway, it wasn't in the first rounds because we all got like different groups of projects. Anyway, we've all got the top 10 and we have to rank them from one to 10, one being the worst, 10 being the best in any various, uh, various categories. Right. So, yeah. Yeah.

Chris Gammell: Documentation and technical ability. That kind of stuff. Yeah.

Dave Jones: How open is, yeah. How open is the product, for example, is a tough one, right? How open is the product? And I've got to, and I feel bad because I've got to rank one of them one out of 10.

Chris Gammell: Oh, no way, man. That is the, that is why you do rank systems. Otherwise you get, you get, uh, you ever done a gauge R&R before? You know what that is? Uh, what? The gauge R&R is like, if you had like four test stands and like three of them were spot on perfect. You're like, sorry, like three calibration stands, four calibration stands, three of them are spot on perfect. And then one of them is just skewed higher. You, the gauge R&R is basically how you, you can, uh, compare all four of those and then just determine, oh, test stand four is actually just higher. So the idea, the tying back to this thing would be, you know, if you and I are judging something and then there's a third person and you and I are centered around five, right? All of our scores are centered around five. Another person centered around eight, you're basically like railing that op amp judging. You know what I mean? Like you're, you're, you're, you have compression of the actual judging numbers. So that's why ranking is great. Full dynamic range.

Dave Jones: Oh, I, I totally agree. Full dynamic range. But in, in some cases, like, as I said, like the open thing, right? They're all open. I mean, that's a definition of the, like, they have to be to get in the contest, right? They've got to be open. So how open are they? Well, like, you know, like there's one or two standouts by, like, I'm, I'm going to judge more, like if you've put more documentation into it, you know, if you've got more documentation, then you're more open in quote marks. So I'm going to rank you higher, right? If you've got more documentation, you're not necessarily more open in the, legal sense, right? Oh, right. You're, you're exactly the same as everyone else. You use the same license, blah, blah, blah. Right. But you're, you're providing more information. So, you know, but yeah, there, there's like five of them in there that are like almost equal. Like, so I've just got to like, yeah, it's tough.

Chris Gammell: Sorry, man. It's tough. Yeah. I know. The life of the guy who's just passing judgment. It's, uh, you know, it's tough.

Dave Jones: Yeah. Anyway. Uh, but it's good. They're all great projects.

Chris Gammell: Oh yeah, man. So I'm always impressed by people doing that stuff. So.

Dave Jones: Yep. Yep. The amount of work they put in, it's fantastic. Some really good ones. Anyway. So I'm going to do that now.

Chris Gammell: Okay, cool. Well, uh, should we call this a success? Is this a success? I think it, well, it worked. Nothing, nothing went wrong.

Dave Jones: Really?

Chris Gammell: Marginal success.

Dave Jones: Woo. Yay.

Chris Gammell: Cool.

Dave Jones: And that's what we do best here on the NPR.

Chris Gammell: That's right.

Dave Jones: Mediocrity. And we are officially in the podcast universe.

Chris Gammell: Oh yeah. I should post that map. That's right. Yeah. Someone made a map of all the different podcasts and stuff. We are on there.

Dave Jones: Yeah. I didn't see it at first. We are buried away. And I, I have influenced nobody else, but you're a, you're an influencer, Chris. You've.

Chris Gammell: Oh yeah. Cause they started the engineering commons and don't do it anymore.

Speaker ?: Yeah.

Dave Jones: So they're, yeah, they, it's, it's a map of how all the podcasts are linked.

Chris Gammell: Yeah.

Dave Jones: You know? So yeah. By the people. Yeah.

Chris Gammell: Cause you can find other podcasts, that kind of thing. Yeah.

Dave Jones: Yeah. It's very handy. Plus link to that. We're on there. Yay.

Chris Gammell: Cool. All right, man. Well, talk to you next week. Good luck with judging. See ya.

Chris Gammell: Bye.

Speaker ?: Bye. Bye. Bye. We'll see you next time.