Parker talks with Brandon Satrom of Particle about the future of IoT and then design and prototype an IoT device.
Agustin Pelaez and Cameron Klotz of Ubidots talk about what is IoT and how to start an IoT Project.
John Adams joins Parker and Stephen to discuss IoT Security, Crappy IoT Devices, and WS2812B LEDs.
Parker is an Electrical Engineer with backgrounds in Embedded System Design and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. The following year he designed and produced an Atari 2600 video mod to allow the Atari to display a crisp, RF fuzz free picture on newer TVs. Over a thousand Atari video mods where produced by Parker from 2006 to 2011 and the mod is still made by other enthusiasts in the Atari community.
In 2006, Parker enrolled at The University of Texas at Austin as a Petroleum Engineer. After realizing electronics was his passion he switched majors in 2007 to Electrical and Computer Engineering. Following his previous background in making the Atari 2600 video mod, Parker decided to take more board layout classes and circuit design classes. Other areas of study include robotics, microcontroller theory and design, FPGA development with VHDL and Verilog, and image and signal processing with DSPs. In 2010, Parker won a Ti sponsored Launchpad programming and design contest that was held by the IEEE CS chapter at the University. Parker graduated with a BS in Electrical and Computer Engineering in the Spring of 2012.
In the Summer of 2012, Parker was hired on as an Electrical Engineer at Dynamic Perception to design and prototype new electronic products. Here, Parker learned about full product development cycles and honed his board layout skills. Seeing the difficulties in managing operations and FCC/CE compliance testing, Parker thought there had to be a better way for small electronic companies to get their product out in customer's hands.
Parker also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Special thanks to whixr over at Tymkrs for the intro and outro!
Hello, and welcome to the macro fab engineering podcast where your hosts Parker, Dolman, and Steven Craig, this is episode 66. Hey, listeners, if you enjoy the map, please let others know about us. Tell your co workers, friends, family loved ones and share it on social media at macro fab, or follow us on Facebook. At some point during the show, we're going to announce a secret code word. If you email us the code word and your address, we will send you cool macro fab swag. The email address is podcast at macro fab.com. Yeah, we've been. We've had actually a lot of people write in, and we've been shipping out koozies left and right now it's been great to hear from you guys. We appreciate you guys doing that. Thank you so much, guys. Yeah. Yeah, if you want to, if you want to just you know, hit us up and tell us some cool stuff. Or tell us what you think about the podcast, please let us know. Yep. podcast at macro fab.com. So last week, we had a our first big meetup or electronics meetup sponsored by Mouser. here in Houston, we had over 90 people show up last Wednesday. And that That's April 26 2017. Yeah, those that are actually counting days. We had some of them share projects, what they're working on. We had a lot of really good networking, just pretty good chill party. It was awesome. For a first time for something like this. We had excellent turnout. Yeah, I think we had like about 120 RSVPs. And 90 showed up that you which is phenomenal. Yeah, yeah. And so I would like to also thank our guest speakers. Damian, who is the president of the Greater Houston Manufacturers Association, Rick, who is the owner and an engineer at sparks engineering, and Ray, who is a professor at Rice in the electrical and computer engineering departments. Thank you so much for coming. And thanks, panel. Steven was actually the moderator for the panel. Thank you, Steven for doing that. Yeah. And I would also like to thank Abel and Iris at macro fab. They're our marketing team. And they basically made all that work. Yeah, they, they, they put a lot of effort into it. It was a lot of fun. And the next one is on May 31. And I think this is just going to be a normal thing. We do the last Wednesday of every month, we're going to have an electronics meetup. So if you are in Houston, or in the surrounding area, come on by May 31. It's six o'clock to eight o'clock. Yep. And, and if you're not in Houston, move here. Okay, on to Stephens design challenge that he put, we talked about. Oh, wow. How long ago? Was that? Like, five years? Yeah. So we've had a string of guests. Yeah. Which is, which for our longer time listeners, is kind of uncommon for us. I mean, we we've had situations where we've had two guests in a row, but this time, we had three. And so it's been a while since we've gotten an update on our, you know, our the projects that we've been are more longer term things that we do that just Parker and I. So, you know, as a lot of listeners are probably getting annoyed with my synth design. I've been working on that which by the way works. It's been five weeks. Unfortunately, you're not doing an update about I'm not doing an update. But everyone just closed the browser. Yeah, right. Yeah, no, done. I'm done with this podcast. Next week, we'll have an update on that. Yeah, yeah. So regardless, it works. But a couple of weeks ago, I gave a design challenge to kind of work on the synth with me. And the design challenge I gave was basically to help create a little circuit that had a ring of LEDs that surrounded a potentially ometer basically the knobs on the front of my synth. I want them to be cool and light up as you turn highly over engineered your super excessive absolute wink engineering. But it was that as a segment wank engineering, the web? I didn't we have we had unnecessary engineering as one segment. That's a little bit more. Like, you know, for work, I think we might have an RFO that talks a little bit about doing so. Stay tuned for that. No, no. Okay, so So the this. Regardless, I gave this design challenge, and we had come up with the idea to just present it to the listeners, like five minutes before a podcast. And surprisingly, or maybe not supply surprisingly, we actually got a good handful of submittals submittals submittals. Yeah, new word. Submissions. I am one to talk about that. Yeah. So so we printed out a handful of the designs, a couple that we chose that seem to be really interesting. and fun. And what's cool is we got a really wide variety. Yeah, we picked it all over the map from extreme digital to extreme analog to off shelf parts. It's pretty cool. Yeah, yeah. Yeah. And, you know, one of the biggest things that I put as a design requirement was that it was cheap. And what's cool is everyone who kind of wrote in, wrote about how their design is cheap, and and why it's cheap. And so we have a variety of costs also. So yeah, let's, let's go ahead and step through a couple of them in and see what we got. So the first one is from Thomas Ives. from Salem, Oregon. Yep. So, so Thomas, decided to go with more of an off the shelf sort of solution. So he was suggesting using a chip called an LM 3914. It's a really old school chip. Yeah. Have you looked up the datasheet? for it? It's definitely been like, Oh, yeah. You're like, oh my gosh, intercell. Have you ever seen Intersil data sheet I've never seen there's so so scan. Like, yeah, but like the lines are like extremely black, but like dirty black. You know, they have like dots every okay. Yeah. Basically, the contrast was cranked way up. Oh, yeah. Yeah. Yeah. Try to get as much information as possible that it picked up all like the skin flakes and stuff. Yeah, exactly. Exactly. And they're there for their sweat. Stains grows coffee, beard hair falling on coffee ring staying on like page eight. Oh, yeah. Yeah. So yeah, this lm 3914 is a bar dot LED driver. So it's effectively a solution for exactly that, Olden? ATS. Probably? No, I think they had these going further back. But still, that's a while ago. Yeah. Yeah. Well, when was the first red LED? First red LEDs? I'm gonna guess the 60s. I'm looking up right now. Yeah, look it up. Keep talking. So Okay, y'all keep going on. So So effectively this is so that you can use this chip as like a VU meter. If you pump a signal into it, it'll actually move a bar driver, because basically, the chip just has a an analog input. So you feed it an analog signal, and it will light up bars based off of a zero to five volt signal. 1907 h j round invented them. This is just off the wiki page. But I would say way, the number that you would actually care about is 1962. Yes, because that's first production. 1907 probably the concept. Yeah, yeah. I was about to say because we weren't doing we weren't messing with silicon devices till the early 50s. Well, they hadn't before that too, I think wasn't the first transistor 55. Well, transistor. Yeah. But they were messing with, with the properties of silicon properties circling in. And the junctions before that. Hmm. We Yeah, I would go with 62. That's the first like, that's pretty cool. You could open up maybe the plumbing out the Radio Shack catalog, but I catalog back then and be like, Oh, I can get you know, a 10 megahertz of an eyes, Crystal and blah, blah, blah, blah. And then there's like, at the very back, there's a whole one page advertisement for like, LED. Yeah. And people are like, No, I bet you they still emits one frequency of light. What? No, I bet you it's one of those things where it's like, it still would like die after like 1000 hours of use or something. It probably wasn't much better than an incandescent, if you know more information about that led, like where we can get one? Yes. Let's buy one just buy one. Yes. And destroy it. Anyways, on on with the design. Okay. Yeah. So so in order to so So basically, we I had a certain number of LEDs. In the design requirement, I think it was 16 to 30. Somewhere in that range, I made it a range. And so by using two of these chips, he could accomplish it by just inputting a, a signal directly to this signal pin. And as you turn the potentiometer, you'd have to use a dual potentiometer on this. But it would it would drive an analog signal and turn the LEDs on. So this is like the one chip package or that you really have to package. And actually one thing is is how would you make them basically be in series because they need to be in series. Basically, one chip needs to go all the way up in terms of the bar graph, the 10 array right, then the next one needs to start going up. Yeah, so I think that there. I mean, I guess you could probably play some games with like a voltage divider on the inputs. Oh, you know what I think I think it was something about I think, yeah, I think there was some kind of voltage divider thing. He doesn't actually have it in his documentation that I have right here. But I'm looking at the datasheet on my phone right now. And I think you can I changed some reference signals in there such that zero to five volts would be one and 2.5 to be these are these chips basically have 10 op amps in comparator mode set. Right, right. That's what it is. Yeah, right. So I think the only thing that kind of gets a little bit strange about this one is that it just dumps a ton of current which is a little bit outside of my design requirements, but a stork still pretty cool. And it's only like two parts plus LEDs, right? You know, so, so So Thomas also suggested a another simple design that uses a pic 12x 675 And Charlie flexing, flexing, you know, Jelly flakes. Oh, yeah. So that's it's a really funky matrixing style, where it relies on tristate Yeah, you tristate pins and also turning pins from inputs to outputs. So you can sync and source, right you have you have sync source, and no. Yeah. So yes. Hi, z. So yeah, that's, that's a cool method to using only four pins, you can drive 12 LEDs, and if you use more, you can drive a significant. Yeah, much larger, Charlie flexing is a little funky. Because it relies a lot on the transistors. And you have to kind of play some games with that. So it's not as elegant. But I liked it. I like the one chip solution that he gave. Cool. So the next one is Brett Smith. So Brett gave us he's from Las Vegas. Bread gave us probably one of the most interesting does that I like this one a lot. So when I first gave the challenge, I was like, everyone's just gonna put a microcontroller and have it you know, spit out pins or do whatever, find some, some kind of solution at Mouser and give it to us. Brett went way off the reservation, not this one, but in a super cool way. So let me see here. i He's got a paragraph here where he's kind of describing the the function. Yeah, it's better just to read that paragraph. Yeah, okay, I'm gonna read this. So actually, you know, in I kind of like it when data sheets have a section that says theory of operation, he basically wrote a theory of operation. Yep. So here's what he said. Basically, an oscillator drives a counter that controls which output is active on a demuxer that drives a buffer for the LEDs. The input value of the demuxer is controlled by a mano stable vibrator 555 with a pulse width that is controlled by the potential ometer. If the pot is turned up higher than the pulse is longer, and more of the LEDs are lit as a result, this leads to the happy coincidence that they leading LED will dim relative to the pots position within its respective 1/16 of the pots 300 degree throw. So wow, he did he did some some. This is almost like a weird intro to digital electronics homework problem. Oh, yeah, in a way where you have to do like, I've got an oscillator, I've got a counter, I've got a D mux. And you have to, you know, you have these like threshold trip points and things. But then he also throws in this buffer chip that, that drives all the LEDs. So it's kind of counts up, based off of your pulse width, it counts to whatever led it needs to and then spits that out effectively. And although that sounds a little bit long winded, and like a lot of stuff, he put the price for every single chip that goes in there. And the total for all of the chips is $1.74. Yeah. Which is crazy. Absolutely crazy. The only thing that might get a little bit crazy on that is the fact that there's 123456 icees that go into that. And but they're inexpensive. They're inexpensive, but the board physical size, I put a constraint of like 1.2 by 1.2. I think you could still do he was 30 mils over. Oh 30 mils. Yeah, he's 1.230 Well, he says we can fit even squeeze it though. Let's Okay, I'll let everyone know a secret when I came up with a dimension sizes. I just shot them out of my ass. I literally just like this big by this big just to see. So yeah, going over by 30 mils. No problem. Really nice. Um, I'll post a picture on the podcast description but Connecticut in Altium in the layout is beautiful. Oh, yeah. Yeah. We have a full page print of his of his layout. And of course, he chooses like a nice blue and an orange so it pops real nice with along with a red. So in fact, a lot of these guys have have gave us images. In fact, Thomas eyes awesome, dude. He drew it up in paint. Way to go Thomas. Way thumbs up. Done. So, so that was a that was a really cool design. And when I first started digging through it, I was like, What the hell is this guy doing? And then you look, it's like, oh my gosh, that's he went definitely on the it's it's really complicated in terms of the engineering. But when you boil down the costs, it's very inexpensive. Yeah, well and requires zero code, zero code. So this is the, this is the analog slash digital version that requires no smarts whatsoever. So you turned on and adjust works. Yep. So next one is David sandal. David, yeah. David sandal, sandal without an E in it. So David actually decided to use a CPL D on his design. So he's, you might consider Dave's to be more of a I guess a brute force traditional design, where you have an ad that talks over I squared C and dumps into a CPL D. And then that just drives LEDs straight from its pins. Yep. Go listen app. Note that that Xilinx? It's a cool runner to CPLD. So I actually looked up the app note that he put in his ex app 805 I guess if you probably Google that you'll pop up. We'll put it down in the podcast description. But the cool runner icon kind of wish that was like an 80s. Movie. As opposed to Cool Runnings. Yeah. The Windy cool runner. Yeah, it's the secret to Blade Runner. Well or Running Man. Oh, running. No, it's it's it's Blade Runner and running them together together. Where it's um, it would be the one of the called Harrison Ford and Arnold Schwarzenegger, our sidekick sidekicks chasing down like a cyberpunk space chasing down replikins? No, it's cyberspace. It's cyberpunk, New York in like 2020. XX is the year. So escape from Cool Runnings to the I'd watch it. Yeah. It was on Netflix on watching. That sounds great. So yeah, no, this is this. You know, if you take a look at the we'll post all these designs, by the way, but but if you take a look at this design, it's it's really traditional, you got voltage regulators CPLD. An add input? Not a lot of magic going on. Magics in the code? Yeah. Yeah, the magic would be in the code in this case. And in fact, what's what's what you could do with this, if you have a lot more control over the brightness, because you could, you could bit bang, you could PWM LEDs. The only thing is it might be you can do a really cool like sequence. So like when you're turning up the knob, it can do like a LED trace, kind of like a Knight Rider effect. Up to the value. Well, yeah, look, or you could do Oh, man, that'd be cool. You can also do, excuse me. Like when you first turn on the device, like they all sweep across and sweep back like it's, quote unquote, calibrating. Yeah. Yeah, exactly. So you know, there's a lot of cool things, you could do a few put smarts into it. In this one, the CPL D only has the capability to drive 8 million per pin. So that's fine. Right now, it's probably bright enough. But it doesn't give an enormous amount of control. Regardless, all said and done, these things don't need to blind you. So ah, probably plenty enough. And then the last one we have on our list is from Sean Swift. We've, we've worked with Sean swift quite a bit. Yeah. And the fab. So it was it was really cool to see Sean. Right in here. So Shawn probably went with the most traditional version of a design here. Probably the absolute cheapest as well. Exactly. Yeah. So he suggested an 80, tiny 88, which is $1. In running is it singles, right? Remember, that wasn't seamless look anywhere? Not sure. But it's cheap. I didn't know Microsoft was that inexpensive and still had 28 I opened? Yeah, so yeah, this thing has 28 IO pins. It has an eight megahertz internal oscillator. So you don't have to have any external stuff. You can almost just put five volts connected the ground and just connect everything else up to LEDs. And you could even tie all the LEDs such that they only have one dropping resistor and then sweep through the code. So it's it's really it's really a one chip solution for $1. So I think Shawn probably wins the cheapest solution out of everyone and he he goes through and you know if we try Plex or multiplexed the LEDs He says you can drive 144 That's right LEDs. So you can probably do your entire panel of knobs. With this one chip. I have another routing on that would be insane. Because you have to cover, you know, was it 19 inches wide by five inches tall, whatever it would be. Yeah, it would be insane. Now, I have 21 knobs. I think 18 of them are going to have the LEDs around them. So 18 times, potentially 30. So okay, that's a lot. Yes, Tommy. This This project is ridiculous. But regardless, yeah, 80 tiny 88 for a buck. Connect LEDs write up two pins. I guess the one of the only things I can think of that is kind of a little bit of a downfall but not really much is that you have you have to rely on driving current through the microcontroller the microcontroller. But that's not that's not a big deal. Not before. Yeah, yeah, it's really not a big deal. But that's the only thing I can think of on this one. That's kind of like, questionable. Well, I actually want to see if a at the internal oscillator of eight megahertz if it can do that big matrix fast enough. So you don't get you know, Blinky LEDs. So So I hooked up an led to our function generator, and I turned it on. I just paused it and adjusted the frequency until I was thought it was all right. Basically, you can't tell. And 55 hertz from my eyes was about the point at which I was like, this is acceptable. And 100 hertz was like, I can't tell it. Yeah, it's anything less than 55. You kind of it's kind of like a It's kind of like a fluorescent bulb. Oh, and when you first turn it on, yeah. And you're like, I can tell that that's blinking, you know. But so 55 was the edge. Now I did some calculations with some other chips. I think maybe I think it was on 16 megahertz. And I could I could drive most of the LEDs for edit like 1000 hertz. So yeah, we actually, when we were developing the macro, watch the pic. This was before it was so on the pic. And we're running the pic at 32 Point, whatever. 30 To kill her. You know, what's the it started? 32 point whatever. time clocks? Yeah, down. Yeah. So we were running off that. And the idea was on the original one was our idea was to make it as cheap as possible. You had all the LEDs that do the time run through one resistor, because you would just like one at a time through a cycle. Well, if you tried to do that it blinked like mad. It was not fast enough to do that. Nice. So I don't. And I tried to make I was like, basically, my code was just straight C. And I was looking at the assembly. I'm like, There's no way I can make this loop faster. So there is a physical limit to how fast it'll go. Yeah. So I kind of wish I need to revisit that project and see how fast that loop was actually running timewise and see if it was under your 5055 Hurt limits. Yeah, part was blank. Pretty bad. Yeah, I bet you it was. So yeah, that. That's that's the kind of designs that we have there. What I think is really cool about this. So we presented four different designs. So the first one was a two chip solution. That was a little bit more of I guess you could say a vintage style chip. Yeah. An off shelf off. Yeah. Off the shelf. It would just work. The let me see here. The second one we talked about that was that Brett? Yeah, we talked about Brett The second time was was a oddly elegant solution using using elegant is a very good way to put it. It was it was really cool. Very interesting. The third one was was more of a traditional, but it uses a CPL D which is different, which you don't see every day. And then the last one was about as traditional as you get just slam on microcontroller and hit pins. Yep. Which Hey, cool. It was when I first presented this, I thought I was going to get a lot of the same stuff. And we got everything very different. Which is super cool. So Brett, if if you want to finish your design out, we'll build it for free. We Okay, so we were kind of discussing this and we both really like your design. Yeah, I think it's super cool. I think because it's just so out there. And yeah, I want to see this built. So yeah, it looks like we have I mean yeah, you got a full schematic here and we have a whole board. So if you want to if you want to get with us podcast at macro fab.com And if you want to build it we will build it for you. and test it out. We'll, we'll pimp it. Alright, so that was the design contest. Yeah, what was the one that you actually designed? So I've designed it I haven't actually built it yet. I do have parts. I have a bag full of parts. But I went with an 80 Tiny 85. And I bit banged some shift registers. All of those dump to LEDs. Gotcha. So, it's, it's a little bit more expensive. So it's kinda in between days when Sean Swift, I was gonna say sharp, swift and Thomas. Yeah, yeah, it's a little bit in between there. And, uh, one of the reasons why I kind of went that direction. Laziness. Like, I already have these parts designed exactly. I have these parts design, and I already have code for bit banging shift registers all burn up. So it makes design turn. You know, I actually do the same thing. Yep. Like, I need a power regulator. What do I already have design? Yeah, exactly. It's like, it's only it's like Parker, it's only for four pads to design a new one. Yeah, I'm like, I'll sit there and design like a whole brand new micro troll. It's got 144 pins. Oh, my God all the time, I think all the time. Like, regulator. Okay, wait, wait, wait with it. Great, great. Side topic right here. When you're designing a microcontroller footprint or pattern or whatever your EDA tool calls it. When it comes down to naming pins. Do you write out the full name from the datasheet. So it's like I Oh, seven slash x tall, three slash blah, blah, blah, an eagle they separate out footprint and symbol. Okay, on the footprint. It's 1-234-567-8910. Both of them are on the schematic of the symbol I do. I write out exactly what the datasheet says I do too. I don't care if it's like 300 characters long. I will write that sucker out. I need to have you do like the next FPGA I have to do Oh god. 1000 pin. Biggest FPGA I've done was like a 288. Pin BGA. But it's on that cuz that's easy. That's like, once you get one row done. It's like copy, paste. Copy Paste. Oh, yeah. But it's when you do the symbol and had an Bert, basically breaking everything out into different IO banks, power, all that stuff. And then doing all the functionalities for the pins? Yeah, yeah, it's, it gets pretty brutal. So I actually the good manufacturers give you a text file that you could parse? Yeah, you just don't dump out the pin num pins, but some don't. Most don't most of the Altera and I think Xilinx both do, yeah, cuz they're kind of they're nice like that. I'll, I wish they could make better IDs. But that's another topic. I actually, so I adopted that method of writing the pins out with all of their functions from a previous job. And we had some really high pin count processors that we were dealing with. And they was so big, it ended up being so big on the schematic that the processor would take up multiple pages. But no, what I really really liked about it, it actually made it look like the processor was physically large. It was like all of our signals are going into this giant monolith into the layout. It's like a 20 millimeter by 20 millimeter part. That's a pretty big part. Yeah, no, that's that's almost an inch. That's about the size of a pic. 32 Yeah, the 144 ish. Yeah, sure. Well, thanks everyone for the design challenge. Those that was really fun. And I think it went well. So we might do more of these in the future. I think we need to do another I think we need to do an official design challenge. Okay, so this was the test. Well, we still picked a winner and but it was kind of impromptu. Well, we actually have a legit one. Like, facing on Mercury. Do you say my challenge is not legit. No. Have our marketing team like actually pimp it out? Yeah, yeah. No, I think it's a good idea. Cool. Okay. RFO RFO rapid fire opinion. So we have three topics. Some of these are a little old. But you know, it's been a while since we've actually had a non guest non guests episode. So pardon if some of these old dusty I ut IoT by Hackaday. Here's why do SAROS press is so expensive by bolt IO. And then Amazon's new echo look has built in camera for selfies. So as the normal thing with our poses, we really hate IoT. And sometimes reason we keep talking about it just won't die. So the IU T IoT by Hackaday is basically the internet of useful things which would actually be i o UT, but whatever Hackaday. Basically Hackaday is because they they run design contests and project contests all the time. is they're looking for useful. IoT devices? Yeah. Yeah. I guess you IoT is probably better. useful things. Oh, okay. Sorry. Instead of Internet have you got to remember, I'm dumb. I wasn't following you. I saw this earlier this week. I think I was out this week. And and I saw it and immediately it was like, oh, Parker read this now. Because we kind of bitch about IoT stuff all the time. So but I brought this one up, because actually at macro fab. That's one thing I've been working on is useful Internet of Things, things, things at our shop. So that we would like to know more data about mine. The first big one is the compressor. It actually was a really good example. What happened on Monday? Oh, yeah. Is I want to basically put an accelerometer on the compressor that because when the compressors turns on, it vibrates like crazy, right? It would be very easy to binary tell if it's running or not. I think that'd be the the code word binary tell binary till. Right us? Yeah, I guess I macro.com. Anyways, back to this thing. Yeah, so accelerometer on it, and then it just connects up the Wi Fi. And then basically, you can know how long the compressor runs for, because that way, you can actually get an hour meter because right, because it doesn't have an hour counter. So we have to change the compressor oil based off time. So preventative maintenance can be based off time, well, no, well, right now it's based off time in terms of like every three months, we need to run time, but runtime is a better way to do it because sometimes it runs more, sometimes it runs less. And then the great thing about that is then you can know when something abnormal happens like a leak in the shop, or like what happened on Monday and a copper pipe that goes to the intercooler brakes and the compressor runs for like two days straight. And I had to go run around all of Houston to try to find the spare parts. And they were way nicer to know that on Saturday. So I could have spent Sunday instead running around finding apart. Right right right. And we probably pre materially aged the compressor by just letting it run non stop stop when the ocean say we didn't let it it had a premature failure. Yep. And yeah, basically the the it has a three quarter inch copper tube, a hard line that comes out of the top of the Compressor Head it's a two stage reciprocating compressor and that goes into a it's not an intercooler I guess it is an intercooler it goes into an intercooler so cools air down before it goes to the main tank. And basically it had a bait why why we pretty much sure what happened is the copper probably work hardened after fiber writing for so long. And then it just broke. And I think what happened is during the move it probably that pipe got because like right on the edge of the compressor. Yeah, and it probably got knocked or whatever and then someone just bent it back. And so because the other one looks fine, right and the other end of that pipe was completely fine. It wasn't work hard at all. And and the the exact area where the copper pipe broke is right at the the stress point of for for vibrating. It wasn't And the crazy thing about this pipe. What is its three quarter inch copper, what is the it's the L type. That's the Elta but it's actually pretty common you can buy this at Home Depot or Lowe's is at the ends, it had a double 45 degree flair. The only tool we could find was a $350 tool at McMaster that has a three quarter inch double 45 flare on the end, right. So us doing it was not in the or us fixing it. Yeah. Was was not going to happen. Yeah. And basically it took me I had to call the company who made the compressor twice. The first time I did it in the morning to see when the spare part and then they had no idea what I was talking about. And then the afternoon I called them again, I got someone else on the line. And I got directed right to like a production manager basically. And he's like, oh, yeah, the part number you have is an old one because we changed everything over a couple months back. So here's the new part number and we have a distributor there in Houston that's got one like actually has that that part. And it was like 430 in the place close at five. It was on the other side of Houston in rush hour traffic and I got there at 502 and they they were nice enough to keep the shop open for me to grab it. So I reinstalled it Monday after work and and that was good. Bob's your uncle So yeah, the in that case, an Internet of useful things would have been really nice. Awesome. I come full circle on that one. Yeah. Okay, topic two. Here's why just sir Is it is it juicero Or Juuso? juicero? You SAROS you Cerro or juicer? juicero juicer? Oh, it depends on the few. Break juice out of that or not. We're trying to say that's one word. So you put this on the list. I don't know what this is to inform me. Okay, so this was a couple of weeks back. There's Oh, no, I know this, this that this is that like juice? It's got like a pre juice bag, and it just squeezes it out. Yeah. No overly expensive mush. Yeah, basically. And so both.io did a really good tear down. Yep. Actually, I really got to tear down of a basic, inexpensive and controversial juicing. Juicing machine. It's a press. It's a delivery mechanism. Yeah, so the best thing is, there's videos of people, quote, unquote, hacking the bags by just squeezing them. And someone commented on one of those videos, I can't remember the guys name. I'm sorry, guys, but the best comment was, I've been hacking Capri Sun since I was five on my hacker, because people were saying, Oh, I hacked the bag by squeezing. That's not a hack. Oh my gosh, we had a conversation about this just the other day that that grinds my gears so hard, bro, like, you know 10 useful life hacks you know these crappy videos on YouTube that show up? Where it's like you're not hacking something you're just using using it in a different way. Oh, is the ice cube? Ice Cube maker? Yeah, they were. Okay, this ice cream an ice cream? It was an ice cube tray and put in your freezer. They put like fruit in in the ice. And so they were like We hacked it? No, you didn't you put fruit in exactly the same way except you made the ice dirty, right? If you melted the plastic down and created something else without it or Yeah, right or like fundamentally changed its use you have now hacked it. You didn't hack it by just putting other crap and you just oh one was okay. There was one that was kind of a hack it they basically use it as a template to make ravioli. It's a tool at that point. Yeah, but then they didn't freeze anything with it. Okay. Its purpose is to put liquid in and freeze well okay. 90% of their quote hacks were still put subsided and the ravioli one was actually pretty good, though. Yeah, I would say that was a hack for an ice cube tray. Maybe? Maybe. But but okay. Yeah, no. So this, this juicer, but quote yet, it has you purchase bags of much fruit. Yeah. And you put it into this device, and click Go. And basically it squeezes the bag. Yeah. So yeah, it just squeezes out and doesn't do anything. Well, I think because basically the press mechanism in his machine is crazy. overbuilt. Oh, yeah. Yeah. Yeah. A lot of aluminum parts. All the gears are actually machine gears are not centered gears. Yeah. Isn't centered centered centered gears which is basically they take powder and powder and resin and lasers and make the gears which sounds more expensive. But it's actually more much more machine. You cook gears, hook gears. We shouldn't mean is that is that hacking gears? Yeah, no, we should have made that the cookie. Code name is cook cooking gears cooking with gears with macro fab. So that's what I was thinking is the reason why that was so builds was I bet you earlier in the development cycle, they were actually planning to squeeze like fruit chunks. And those bags. Oh, so you a would actually squeeze like they'll have like chop peach or whatever in there. Instead of puree instead parade and then a would squeeze it. And that probably just didn't work out. And instead of redesigning it, they just said ship it. Well, you have to remember what the original price tag was. It's like $700 Yes. 699 for a squeezer. So after this got all kind of released and people flipped out. It's now dropped. It's like $350 Hmm. So but and the other comment I like is what people have been saying is it's beautifully engineered. No, it's beautifully made. Poorly engineered because an engineer what it like made like roller like cheap rollers and it just like rolled out like a polar toothpaste, toothpaste or I was thinking Polaroid but I actually think only probably half our listeners will know how a Polaroid were Yeah, yeah, I think I think you're kind of right on that. The engineering behind it, although it still could be beautifully engineered if the engineer was forced to design to those specifications. It's also if they were planning on the price point is $700. It's like it's like a Ferrari or a or a Mercedes a higher end Mercedes because you can buy an expensive Mercedes brand new now. Yeah, but like high end ones were this like, there's some subsystems that are overly complicated like a lot of I think it's like mid like 60s era Mercedes have like, these insane vacuum and hydraulic control systems for everything like door locks are pneumatically actuated. For some reason, I can't remember that's Mercedes when I but I, there's some complexity for complexity sake or sake. Yeah. And if that was their grocery cars tend to do that. If that was their goal, they they achieved it. Okay, if they originally expected a $700 price tag, they probably just sat down and just went wild with the design because they're like, We can do anything for $700. You know, do you want to use the the best plastic? Sure, use that? Why not? No. Do you want to mill our own gears? Sure. Why design our own power supply design or have our own motor? Yeah, everything in there's custom. There's no more shelf stuff for a bag squeezer. Squeezy. topic three, Amazon's new echo look has built in camera for style selfies. So the idea behind this is the Amazon Echo is a device that controls your home, or allows you to order stuff or turn on music, blah, blah. But now it's got a camera and an app on your phone. So you can see how you look through the Amazon Echo. And I'm like, do these people not have mirrors? vanity.io? Yeah. But it will also make suggestions. I think I think what they're trying to do is it will make suggestions on what you should wear and buy. Oh, the Amazon store. Yeah. Oh, yeah, kidding. Yeah. I don't know if that's implemented yet. But that's, I think that's where they're going. Because then if you have a picture of you standing there with clothes on, it can. What if it came with doesn't do this spotlight? If it came with, like, clothes that were like had a pattern on them? And so it would it could figure out the pattern and then basically Photoshop different outfits on you. Like augmented reality? Yeah, see how you would look just overlay? Yeah. But because you're not gonna stand perfectly straight. And you might want to see how a side profile is a good figure that out by the pattern. profile looks terrible. My Profile me sideways is bigger than then. Alfred Hitchcock kind of side. But yeah, like kind of like how a QR code will give a marker for like augmented reality? Yeah, the same way if you can wear like a shirt that's got a marking on it. And so he could figure out an overlay, like a 3d model of the that'd be really cool. Yeah. But that would be that there would be some serious software behind that. Oh, yeah. The hard core software. Yeah. And then it's also like, if you know, parents, it could be because I think the echoes out how also has like, parental control, you could control what your kids wear. So they can't go to work. So you could you could dress them by just like saying you're wearing this, this and this today. Yeah. If you could upload whatever their clothes are into it. Yep. Oh, that's kind of creepy. A little bit. It's like, you're not gonna wear those ratty jeans. You have to wear you know, Sunday's best or whatever. If I had one of these things. And if it could talk to me, it would be like you have worn t shirt jeans and boots for 65 days straight. Do you want to wear something else? No, like, No, thank you. No. I live in Houston. That's every day. It doesn't get cold enough. Right. And it gets too hot to wear long sleeve shirts. Right? Just wear short sleeve jeans. That's just the Houston look. Yeah, right. Yeah, no, it's just mine would be like I give up. Life is not worthless. So I think that will wrap up this episode of the macro engineering podcast. It was episode number 66. We were hosts Parker Dolan and Steven Craig. Let everyone take it easy.
John Adams joins Parker and Stephen to discuss IoT Security, Crappy IoT Devices, and WS2812B LEDs.
Parker talks with Brandon Satrom of Particle about the future of IoT and then design and prototype an IoT device.
Agustin Pelaez and Cameron Klotz of Ubidots talk about what is IoT and how to start an IoT Project.