This week's topics are: Porsche's Synthetic Gasoline, Record Chip Manufacturing Sales for the year 2022, and the Raspberry_Pi Social Media Firestorm.
What are the common Design Rule Check errors that Parker and Stephen see as Contract Manufacturers? Are these DRC errors the ones that you run into?
PCB serial numbering? Parker and Stephen cover their thoughts on applying a unique identifier to PCBs in production for inventory and testing control.
Have an idea for a podcast episode? Maybe a question for Parker and Stephen? [Hit us up!](mailto: podcast@macrofab.com)
Did you enjoy this episode? Let us know in the comments below!
Figure 1: SAIM software test.
Figure 2: Veroboard Stephen is using.
Figure 3: Stephen’s speaker kit under construction.
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. We're your hosts, Parker Dolan
and Steven Craig.
This is episode 76.
Hey, listeners, if you enjoyed the Mac fab engineering podcast, please let others know about this. Tell your co workers, your friends, your family loved ones, share it on social media at macro fab, or follow us on Facebook, our Instagram is macro fab Inc.
and our like thank you, our listener, or if there's multiple people listening to the same set of speaker listeners for downloading our show. If you're not subscribed to the podcast, go and click the subscribe button. And that way you can get the you know, show right when it comes out on Friday at 11 o'clock
episodes. Yep. If you're on iTunes, please take some time to submit a quick review. It helps the show stay viable on iTunes, any visible neighbor, this is the first time that we actually do an edit on this thing. No, we're not doing that. No, no edit. Okay, we're gonna be visible on iTunes. If you guys do a review for us, please review us on iTunes and and helps new listeners find
is there actually any reviews on iTunes? So, first, we
may have one review on iTunes, please go and
click five stars and type in first. Yeah, and hit Submit first.
So yeah, at some point during the show, we're gonna have a code word. If if you email us this code word and your address, we'll send some cool swag your way. The email address is podcast at macro fab.com.
Yep. Great. We got that done. Finally. So we have an update for a previous podcast the and not XOR guys that did the DEF CON 25 the bender on a bender. Yep. And they're on Bender DEF CON badge. They had a really cool Hackaday article detailing the features of the badge and well, only some of the features only some of the features that are they were allowed to show,
right.
And when this episode comes out a little bit later on was it'll be July the 14th. Around 11 o'clock central time, that's when the subject comes out. There we'll be doing a hack chat with Hackaday. So go check that out. Yeah, that'll be
cool. Talking about all those all those cool badges.
Yes. Lots of lots. There's a lot of badges out there. Yeah,
there's a ton and we built a bunch of them and programmed up on a bunch of them a whole bunch. So yeah, though they the the bender on a bender badges is really cool. And that was cool to see Hackaday do an article on it. I'm really interested to see what what's gonna come of that, because we had them on as a guest a few episodes back, and they kind of released a little bit of information. There's a little bit more. I'm curious to see how deep this badge goes. Yeah. How deep is the bender hole? Yeah, yeah. And Parker and I started talking about like, if we made our own badge, like, what are some secrets that we could add into it? Yeah, so that might be something we do in the future. Yeah, for some fun.
So last week, we had Ben and Chris on, right. And we talked a little bit about the Raspberry Pi compute module LVTs test board, and how that all works. Yeah, I still haven't received my six pack of beer from from Steven yet.
I just come in, I just have to pick, I have to find the worst possible beer that I can buy for you and get you that so I can fulfill my part. I think I've selected a good candidate.
Okay. So, having that done, basically, that hardware done, I started basically grafting on to the current pin hex system. So pin hack, Rev. Eight is in full swing at the moment. So hopefully in the next month, I get that built in prototyped out and that all works.
That's that board has a lot going on. And you're taking the heart and soul out of it. And putting a little brown in the heart or soul. Well, okay, yeah. So you're right, you're right, the the AV heart and soul, you're taking that part out of it, and you're slapping in a much bigger heart and soul into it. So that's gonna take a while, right?
Yeah, it's well, routing wise, it won't be too bad. Yeah, it's because it's not actually a lot of no signals throughout. It's mainly making sure that the power is gonna be clean enough for that Raspberry Pi three,
well, I mean, does the Raspberry Pi like how do you know how much ripple it can
withstand? Well, it's actually a you you basically are supplying the actual BGA chip directly With
the voltage, oh, it's got to have some bypass caps on there, it
doesn't really have too many bypass gates really only has enough for just the local area. Okay, so you have to make sure you give it clean power. Oh, I don't know if there is a ripple requirement though. But I made sure to get pretty clean power in it.
You know what those, those little boards? I haven't seen many ripple requirements that kind of just like, give it some power,
give it you know, 1.8 volts, and it'd be happy. Okay, so
you got to stack some caps up. Right where it connects to the Yes, yeah. To the your board. Cool.
Well, I was going to put actually, it's going to have its own basically sub power circuitry that will feed off the 12 volt line.
Oh, regulator just for it. Yeah. Oh, that's a good idea. Yeah. That way, it's more localized. And, and it'll have its own separate return path for ground. Yep. Okay. Yeah. Well, that's that's probably the best way to guarantee
that that part works, right. Yeah. And there's nothing really on further on that side of the board. That's like, super critical that it's good.
Does the cars run on five volt?
Five 2.5 3.3 and 1.8 volts.
Okay. Yeah, I think you mentioned that last week, but you have to give it every single one of us. Oh, so you have to make a string of regulator. Oh, yeah. That all feeds it. Okay. Okay. So we I really over that's how they get the cost down. You have to give it all the power. Yeah. That's it.
I, on my, my board. I basically said like, I put one family of LTO. And just like, every single Yeah, all the ones in the spectrum. Yeah, well, that way I was, you know, on enemy one footprint. And so for a prototype that works. Because you don't really have to do a lot cost savings for a prototype or for not really a prototype, it's more of a proof of concept to the hardware. But for the pen hack board, you know, every you know, dollar counts if we can knock it off. And so, on that, that board I actually put jumpers that the you can unplug that breaks the power line. And so you can put your meter on it and read how much amperage that line is pulling. And so I can spec each regulator for that power, plus a margin a bit. Gotcha, gotcha. Actually, that's a good question for for our listeners is because I don't know it is how much over spec should you spec your lD Oh, four.
In terms of its maximum? Yeah. So
let's say your average power consumption is 100 milliamps at 3.3 volts. Do you go 125 milliamps is what you need your LTO four, do you do double or?
Ah? Okay, so here, I can answer at least a part of that. Okay. So, let's Okay, let's just say you have five volts coming in, and it's, you're spitting out three, three, yeah, at 100 milliamps. So subtract the two. And that's how much that the regulator has to burn. Yep. Right. So you got what does that one? seven volts. So 1.7, at 100 milliamps. That's how much power it's going to burn as just heat. Yep. So what you have to do is then calculate what is the temperature rise of the junction in the regulator, basically, well, that's that's your package size, right? And that will determine your package size, which will then determine your heatsink size. And that is the point where you have to make the judgment call on what temperature rise you are comfortable with.
Oh, I'm talking about like, just the amperage rating of LDL.
Well, okay, so what is the Okay, so that is is that 100 milliamps? So let's say you draw
100 milliamps like, do you need to spec 150? Is that like, or can you spec 100?
Well, okay, it all depends on on how stringent you want to be with it. Does the Raspberry Pi compute module pull a constant amount of power? Or is it spikes? Or Yeah, well, spikes are one thing. But is it? You know, is there a situation where for some period of time it's going to demand 200? Yeah, or something? So it depends on on your however you design works. In that case, I would think that the Raspberry Pi is pretty fixed with its current demand, unless you hook it up to something that asks for more.
I'm not real. It's actually really good at throttling its processor. Yeah. And so it actually doesn't take too much power. One's just sitting there. Yeah. But when you're like, decide to compile some code on it, it cranks up pretty good. It'll rip.
Yeah. So actually, kind of on a side note, I found something interesting. I don't, probably cuz I don't know too much about the Raspberry Pi. But I have Raspberry Pi's working for my scanners on our manufacturing for barcode scanners. We use Raspberry Pi's to report back to our system because it just works out well and they were cheap. However, in the specifications for the Raspberry Pi, I think it says something Like the USB can supply 100 milliamps of current tops. Yeah. So I, the scanners, I plugged into these Raspberry Pi's operate at 400 milliamps. So I was thinking that I needed external power for my scanners. But I just said, you know, screw it, and I plugged the scanner right into the Raspberry Pi, and it works fine. It's taking 400 milliamps through the USB ports on the Raspberry Pi. And it's been doing that for months,
it might be 100 milliamps per port up to a certain maximum, and you're just pulling the maximum through one port. You
know what? Well, yeah, you're right. There are four, there's four. And I'm only using one. Yeah. So that Yeah, that might be it. It might be fused for a little bit over 400. That makes sense. But I think I read the I read somewhere in some kind of requirements document that it was per port, it was 100. I don't know, whatever, what it I did the empirical thing where I just plugged it in at work. So I said, Whatever screw external power. So and then,
two weeks ago, I was talking about the Python open CV stuff. And I finally got phase zero of our, like one of the longest running like projects we've had on the show the same machine.
Okay, it's the same machine goes way before that way before the show. Yeah, it goes so far back that, in my interview to get a job at macro fab, I talked with Parker and Chris about designing this. Yes. So what have you accomplished now? Parker? So
right now, I, I basically had I taught myself Python, open CV. And 3d printed up a chassis, put a webcam on it, and it looks down at a board, right. And so you put a board in it, it takes a picture. It does Lens Correction, because it's all like, you know, fish died, right? And then it's automatically it basically loads up our Makram API, pulls all the information down, and then crops the parts out. And that all works.
Right. So exciting. So you the interface on on our system website, you go and you see your board. And it has an XY RS file in there that has the XY location of our chart. Yeah. So you can put a board under your camera, and it will take a picture of the whole board, but then crop out based off of the information on our website, a picture of each component. Yes, that
is awesome. Yeah. And today was like the was the was it the
red letter day? Yes, I'm
like, something like that. Whatever that means. Sounds like it should be something that we should be doing.
Yeah, right, saying. So originally, this device was supposed to be we're not supposed to be there's been talk in the past about this being on a on an XY gantry, like a CNC where it would just drive a camera over apart, take a picture based off of its X, Y or s data. But in your case, you're just taking a wide field image and then cropping it based off of each component. Yeah,
the problem with driving the camera on the each part, what we found out is you need I've really, really hos gantry system to move a camera fast enough. So that would work well. So it's viable. Yes. And so this is faster, basically, because it's just software, chopping the image up, now you get into issues where like, if the board, like, towards the, like left side of the board or towards the edges, like you start seeing more of the side of the part than the top of the part. Yeah, because how the parallaxing issues of the lens happens, right. So we're probably going to move to a kind of a gantry system where basically the camera can, because right now only does one board, but won't be able to take a picture of a whole panel, right, so you'll pop the panel in, and the gantry will move the camera over one PCB, take picture of the whole PCB picture, it can have that yeah, and basically just stitch it all together in, in the software. And and
so the kind of one of the original goals of this project was for it to be kind of a QC check thing, yes, where an operator instead of having something under a microscope, which will still do for for certain situations, but but an operator can look at images on a on a, on a computer screen, and and do quality control based off of those kinds of images. And if they need a microscope to look further into things, so it's nifty. It's really cool.
And so the next step is to basically up the camera quality to something that's not you know, a potato.
Well, the camera quality determines what there's a lot of things that determine The quality of it but I mean, we need to be able to look at Oh 402 components, any on smaller 202 a 1021. And with enough resolution that we can inspect the solder joints on an OH 201 from an image taken inches above the board. Yeah, so it's got to be a badass camera. Yeah.
So I've been looking at some, and I'll have more information about that on the next podcast. So if anyone out there has any recommendations for basically a setup that's like this, I'm looking at, like, you know, 1516 megapixel USB cameras to do this.
Right. Yeah. Right. So and better is better. Let's do that. Yeah. So get get it get a I mean, in other words, like, we're willing to invest in a good camera.
Yeah. At this point. Yeah. The, I'm thinking about making like the view area around two inches by two inches. And it's have that go to across the whole panel.
What does it do to Okay, I just wanted to make sure Yeah. noise that's noise. I know what everyone's gonna know what that means. Yeah.
Steven, what do you been doing
a handful of things always doing a bunch of random crap. So I got a notification the other day that my my synthesizer parts finally shipped after ordering them like a month ago. But it was one of those things where it's an obsolete chip I ordered from from a place that says they had it, they didn't have it. So they had to buy it from somewhere. Probably eBay.
Or you could have gone even
directly. Yeah, no, I could have gotten the E bay, but they the CIT they had they said they had stock and it was 94 cents a chip. I guarantee you I bought five. Okay, and we're shipping it came up to $9 I guarantee you they had to go to eBay and pay like $9 per chip to fulfill their order. So hey, whatever, they're the ones that said that they had quantity in stock. So
regardless, that was like an old listing that they just never removed off their website.
Probably because they're the the only person on earth who has that and they were like yeah, we've got plenty in stock. Well, I mean, I guess they shot themselves in the foot. And you know, if they sent me an email and we're like, saying, oh, sorry, we don't have this in stock can we cancel your order? I probably would have been like yeah, sure I'll just go buy it off ebay but they they were literally like no, we'll go get it for you. Okay, fine. I'm okay with this. Although the problem is I have no idea where they're getting it from so it could just be garbage.
They know the Craig name oh, that's it sound Yeah, right. They
don't want me to blast them on. Yeah, cuz I haven't said who I ordered it from. Yeah, right. So yeah, that that'll be coming in soon and I can get back to making the filter Yeah. So actually, I was working on the filter the other day just getting jumpers on my board. So I you know, I bought one not bought I grabbed from macro fab one of the cases of you know, everyone has them those jumper all breadboard. Yeah. And I like it because I have a whole bunch of different jumpers on the board that I'm working on. And so it's like, rainbow colored with all these jumpers all over it is pretty cool. But I've been I've been doing these on various board or strip board. You know, I've never used those before. He hadn't. Okay, so I was gonna ask you that. I've built a bunch of strip board stuff.
I've only done like the just the single point. I knew wired wire.
I hate those.
I did it all the time.
Well, okay, I've done a bunch to I just can't stand having to like, either try to drag solder from point to point. Or you have to cut like a strip of wire and then kind of like otter in a line. I just
did. I took like the old style. Hard drive cable. Yeah. And then that's, you know, that's actually interesting thing is in the future. Hackers won't have hard drive cables to make inexpensive like hookup wire. Yeah, cuz it's Saturday. SATA cable is terrible for that kind of stuff.
Oh, it's really Yeah, it's really thin. And yeah, carry anything. Yeah.
Yeah, I just take like, old style hard drive cable and make jumpers that like old
like, old like flex scuzzy drive stuff. Yeah, yeah.
Well, IDE is what it's called. Okay. Yeah. I should take a picture. I think I've got a dev board. My first dev board ever built with a parallax propeller with huge? No, no, it actually looks really good because it's made of the perfboard Yeah, single, single, you know, dot perfboard. And then I put it on a aluminum backer. And so it looks really nice, but if you opened it up, it'd be like a rat's nest.
Of course. Yeah. Well, that's the whole point. Make it look nice on the outside. It can be ugly as hell on the inside, right? Oh, yeah. But okay, so So Vera board. For those who don't know It's basically think of a breadboard that has an array of holes in it that are point one inch apart from each other. And the the strips running horizontally are all connected together, not vertically, they're not connected. All you do to create a circuit, you place components in in the various board and on the backside of the board, the copper clad side, you just run a drill bit on these kind of, you know, horizontally connected strips, and you break the traces to make connection points. It's incredibly simple. It's really fast. And it ends up making prototyping. It's kind of like the the the logical step for prototyping. You start off with a breadboard. Then you take the breadboard, you go to something like their board,
because verb board is a simulation of the breadboard in terms of the layout,
yeah, effectively. And then after varial board, you go to an actual PCB, but their board in most cases is enough by itself to get simple circuits done. Like the filter I'm doing. It didn't really warrant going and making a whole PCB for it. However, if I did go make a PCB for it. I would have had it in time already been built? Yeah, it would have already been Yeah, it would have been built. So in this case, I was trying to do it to make it faster. But waiting on parts was kind of goofy. Regardless, I've built a bunch of audio circuits on their boards, and they're fun. And I bought some the other day from a local electronic store and Ace.
No EPO,
I, you know, I don't remember exactly where it
was for fries, three posts. In other words, I
had some Radio Shack doesn't exist anymore. But they were not fantastic quality. And the copper was kind of crappy. And they were sort of already corroded by the time I got them. And I got a sign
that that PCB hackers don't buy stuff anymore. Maybe we will just buy stuff online. And well,
that's just the thing. I was just about to say I bought a board off of Mouser. And it's awesome quality. Yeah, it's super good. And it's like real thick copper. And the thing is that the stuff I bought locally, the holes were not centered on the traces and stuff. And they were all like they felt like different sizes. On top of that it didn't have mounting holes pre drilled into the cost difference. Oh, maybe $1. I mean, okay, so locally, it was probably $4 and Mouser was probably five. But the Mouser has this very board, the part number is 854 dash S T two, it's 80 millimeters by 100 millimeters. So you can create nearly any simple circuit on something that big. So I bought three of them. And I'm going to stack both my envelopes and my filter. Oh, and these. I'm just going to put standoffs in between all of them and stack that inside of my case. And then the wires can just exit from there. The rat nest. Oh, it's absolute rat nest because each board probably has 10 wires coming off of it. And some signals some power. It's gonna be garbage. But regardless, I'm really happy with it. It looks pretty cool.
That's gonna be the magic. The secret sauce for the synth is all that you know what's gonna happen every board
is very secret sauce. Because if you look
at the code word secret sauce, secret
sauce, okay, what's the code word? Yeah,
so the Yeah, it's what's gonna happen is you're gonna get it all together. It's gonna sound like just how you want it. Yeah. And then you're gonna update your design it to make a new PCB. And it's not gonna sound like that.
Oh, my God. Okay. I cannot tell you how many times that's actually been the case. I've done that. And been so frustrated before. Where this piece of shit? rat's nest sounds awesome. And then like a highly engineered device sounds terrible. I've had that happen more than once. And I hate it isn't your what tell tell about your your cardboard.
It's like a preamp. Yeah, just that's a cardboard. No. Okay, so get this. That's like the best thing I've ever seen in my life, by the way.
Oh, it's it's, it's horrendous, but it's awesome. Okay, so
honestly, awesome. So,
you know, and a lot of our listeners know that I've done a lot of work with guitar amplifiers. And the thing about guitar amps that it's it's interesting, especially the tube guitar amp world is that the difference between each guitar NPC out there really isn't a lot. There's not a lot of difference between them. Maybe a few component swaps or component changes here and there. But not a whole lot. So what I did was I ended up getting some perfboard not perfboard cardboard
what's it called? Cardstock cardstock. Like like
what you would see an old them in a high school science it took a fair
box of frosted Mini Wheats. and basically cut it open. So what
I did was like cut this big piece of that. And on the backside of it, I made a generic preamp, and it's a preamp where every single component in the entire preamp is variable. So if there's a resistor, I put a potentially ometer. If there was a capacitor, I put a rotary switch with like 10, or 12 Different cat values on there, it looks like tumors coming off. And it's an actually, if the caps are small enough, I put variable caps in there. So basically, it's a preamp where you could change every single value on the fly. And it looks horrendous. This thing because I mean, it's like, it was like, cut with scissors. And I use the Sharpie on the front of it to draw the schematic
of it. And it's got like macaroni art on it. Almost,
it looks that bad. But it sounds awesome. This thing sounds incredible. Like I could never, it could never be a product, it looks like garbage. And I guarantee you all the Voodoo behind that would go if I turned it into a PCB, it would probably just every setting would sound like crap. Yeah, you know,
where you That's what you said is like, there's some settings that sound like amazing. And then you click it one more and just like,
oh, yeah, yeah, well, I mean that. And that's just the thing, like, when it comes to this is an interesting point on design. Engineers have the entire sandbox available to us, right? You know, we could, if we design a product, we could allow a user to modify anything, right. But most of the time, design work is less about actually designing a circuit, it's more about restricting what the user can do. So on a guitar amp, you you, you kind of put someone in a box, and you're like, This is the sounds you can play with, I'm not going to let you go outside of these boundaries, and whatnot. But on this preamp I built, I was like, you can do anything you want with this. It just turned knobs. And what that allows for is equally as many wonderful sounding things, and absolutely horrendous. And it's fun to play with.
So yeah, we got a good picture of that thing.
You know, I gotta go resurrected. It's in a laundry bin in my old shop, right? I even had it to where I mounted the tube sockets to the cardstock. So you literally have to jam a tube into cardstock. In order for this. I need to bring it in. Awesome. So yeah. So I was playing with that. And so the the synth parts are the I don't know, I got an email saying they're supposed to be here. I checked my mail yesterday. They're not here. We'll see. We'll hear when you get home tonight. Maybe maybe we'll see. So I'm going to add just a side note, I'm going to add a tube preamp to the synthesizer. And this is already well, yeah, a little bit, a little bit. But the good thing about this project is it doesn't require any work. Because I've already done all the work years ago, I built a two preamp on a single PCB. And it only runs on 12 volts. I used a 555 timer and I created a like a ridiculous step up. Switch mode. Power supply on there. Oh, so 12 volts goes up to 350 Oh, on on a on an OSHPark board on a little OSHPark board again. 350 volts, works great. So all I need to do is just plug it into the synth and I now have a two preamp on there. So I'll be adding that. And then
we should post the circuit example for that 5.5 timer voltage. You know I would that be I sent a coupler. So now hired well
to 350 it's
a lot higher up I was 59. That's times 30
or something like that. So
it's it's actually 29.166 repeating the course of course, yeah, of course.
So, yeah, I kind of a little bit found lifted that circuit, I found an example of that. And I modified it just a little bit. And works great powers the whole thing. So and then the last thing is we finished the space echo a few weeks ago. Yes, it's sitting right over there, which means we accomplished it's a one of the first if not the first podcast projects that we've been working on. So I actually kicked up an old podcast project, which was the new tube amp. Oh yeah. So I started building the tri Trix speaker kit from parts Express. I think that's the name of the place. So I've got a pair of hoity toity Hi Fi speakers that I'm building at the moment to go along with the new tube amp.
So that new tube amp we have we actually never said this idea on the podcast but always wanted to do it. So the new tube amp for those that don't know I'm pretty sure we talked about it. No, that's not not the amp. This other idea.
Oh Oh, yeah,
I know the audio. So okay. Yeah, the new tube tube. Right. I guess it's basically a dip socketed vacuum tube. For those that have probably noticed CFD package. Yeah, yeah. Well, vacuum fluorescent display style package.
Right. Yeah. Right, right.
And so we have this idea because this is a really compact to basically runs on really low power. Yeah. Is to basically make a add on for your phone. Right. So it's like a like a. A case, I guess, that you slap your phone into, and your phone. headphone. output goes through a headphone preamp. That's a tube.
It's an actual vacuum tube preamp. Yes. For your phone for your
phone. And then Apple decided to remove the 3.5 millimeter
audio jack. Right. Right. So we kind of got screwed on that.
But Android still supports it. Yeah, that's right. Well make an android version.
Right. So for the for the for the most cork sniffing audio file. guys out there who would want the purest of tones, which, if you're running through this tube, you're not going to get the pierce the tones. But regardless, if you want to be the guy who carries around a tube on your phone, we were talking about making that
actually would be really interesting is if we could also combine it with a speaker. Hmm. And so you hold the phone is like a tube amp
box, all powered by the phone. You can
do it by the phone or just have an external battery to
take the vote. So big matter? Yeah, no, but I think it'd be super cool. And we had we had planned on doing that, actually, on top of that the new tubes are actually two amplifying elements inside of one. So you could do stereo with a single tube. Yeah, so we need two speakers. Yeah, you're right. You're right. So yeah, I mean, so the thing is, I have the board built for the actual new tube amp, which is a Hi Fi stereo tube amp, I have the board, I have the transformers. I've got almost everything, except for the enclosure. Really, that's this, that's one of the other than this good pair of speakers to plug it into. That's, that's all that's left on this. So I'm building the speakers right now. The actual enclosures for the speakers have been built, I just got to paint them and finish them. And Parker, with his new 3d printer, actually printed out some plates that I found on Thingiverse, somebody had actually created these plates that hold the crossover components for the speakers on there. So we printed out a pair of those. And it just basically you glue, glue the
inductors and capacitors on right and kind of solder them together.
So it's it's not like amazing, or anything like that. But it does make the management of the crossover look a lot better than you know, most people when it comes to crossovers. They'll just take a piece of MDF and hot glue components to it. But what this does is it kind of manages the inductor. So that's really what matters. You want to make sure the inductors are, you know, 90 degrees out of phase with each other or you'll pick up things well, there's
that term die axially non die axle. I'm
sorry, non die actually alive. That's right. Yeah. So yeah, that's that's
what I've been working on. We're getting an email from some gray beard saying that's actually a real term or something. So we'll go on to the power Pick of the Week. Yep. So this week, we have the Nova tone, or Nova tun, and au 7802. And I found that on the nice chips subreddit, that's reddit.com/are/nice chips. So this is the rundown of that chip. It's a precision low power 24 bit ADC, which is that does exist, right? Has an onboard low noise PGA and normal Yep, onboard oscillator. That's cool. Precision sigma delta ADC. This is where it gets interesting. It's capable up to actually 23 bits of a knob or effective number of bits performance.
Okay, so enough, I've got I've got a quick story about that first job that I that I worked at, we were doing some sensing applications and on one of the boards that we had we had a 24 bit add and you know, I'm I'm still kind of fresh to the game here. So I didn't know what he knob meant. And my boss's saying, talking, like just going on and on about 24 bit add. He's like, Oh, yeah, so this ad DS got 22 e knobs. And I was like, Oh yeah, Yeah, that's pretty good. Yeah, that's that's really great. And then, luckily, like, that ended up being like, true. That was actually okay. Like, I left that meeting and I ran right back to my deck. I was like, what is enough? I was like, oh, geez, that's, uh, that was embarrassing. Because like, what if I was wrong in that situation? Yeah. So. So but 23 bit enough is ridiculously basically,
your least significant bit is just flipping around not doing anything. Yeah. The interesting thing is that comes in most of these like, really crazy high performance. ADCs come in, like, really? Fine Pitch packages. Yeah. This comes in the SOP, 16 or a dip six, or dip? 16? Can you? I wonder how many bits you lose. If you put this in a socket?
You will lose. Okay, so here's the thing, actually. So what's what's funny about this is yes, 23 Enough is incredible. That's 23 from the chip. And just as a little bit further enough, effective number of bits is how low this thing will actually accurately read. Yeah, so you know, if you got 23 bit ADC, right, right, you you forfeit the last the 24th bit, which the 24th bit on a five volt system, or even a three, four 3.3 volt system is like nano volts, so you don't really care. But here's the thing. A knob is mostly based off of your PCB layout and not the chip itself. I mean, the chip in the most perfect situation will get 23. But if your layouts even slightly crappy, you'll get like 21. Yeah, or something like that.
And it communicates i square C, which is kind of cool. Mm hmm. And it's only $2.22 in signal. So that's
insane. Yeah. Because 24 bit a to do with this kind of stuff, especially the fact that as a PGA. Which is a programmable gain amplifier. Yep. And it does all the all of its work on board. Yep, that
is awesome. So the gotcha. Oh, so to get that precision and all that stuff. 10 samples a second, and you only get one game, when next game. When you start gaining it up, you lose the 23 effective bits.
It's like it's like a Greek tragedy, like, all this amazing stuff going on. But like there's something that just has to like, kill you. Yeah, something like that. Push the boulder up the hill, and then it rolls down or something like that. But to be honest, 10 samples a second at a PGA of one is not that bad. If you're doing something like measuring temperature. Yeah, if you're measuring temperature and you want unbelievable. Oh, yeah,
it got awesome. You know, it's got the built in sigma, Delta, etc. So yeah, you can read a thermocouple directly. Yeah.
But but take for example, here, go figure. I'm going to talk about an audio example here. But but if you wanted 192, kilohertz sampling, 24 bit sampling, you know, that chip would be $80 for, you know, in singles and things, but 222. And there's a ton of applications where monitoring 10 times a second is
way more than an F temperature and stuff like, yeah,
they stuff that changes slowly or something that you like, when you press a button. You need a reading right then yeah, that's perfect. Yep. For 222 Not bad. I could see a lot of Arduino shields using this.
Oh, yeah. Especially with it in the top 16 package. Yeah, yeah, for sure. So the RFO Yep. This is one that you found earlier this week. Yep. It's at last or almost a cell phone with no batteries. This is crazy on Hackaday. I don't think it's as crazy as you think. And into that. Yeah. And then the second RFO is evaluation boards for USB type C power delivery. Electronics weekly. We talked about USB type C a lot on the show, almost as much as IoT. Well, how crappy I will and
we talk positively about? Yes, I see a negative effect. Yes. Almost universally.
Yep. Okay, so the cell phone with no batteries. This is cool. It's research by the University of Washington. The tagline that they put out is the first ever battery free cell phone able to make calls by scavenging amiot power. It's not a cell phone. Right? Yeah, it's a it's a remote handset that talks to a custom receiver. No, no, it's impressive what it does, yes. Is it scavenges power via the radio waves going on? And how it transmits the, your your voice, is it? It doesn't encode it into a digital signal? Right? Yeah. It uses the vibrations and some know beyond a paywall. I couldn't go past because I couldn't you know, see the actual paper it
right this is this is incredibly academic right now it's worth segment and in this article was incredibly click Beatty over this was this was the BuzzFeed of Hackaday Oh yeah, for sure. Because I it's not your it's not a cell phone, it does not connect to any cells, it doesn't do anything of that sort. In and it's it doesn't work fantastic, but it works. Yes. And really what I mean it, it basically sniffs power out of the air because, you know, radio and every other whatever that's floating in the air. EM waves, we just blast a bunch of power out. So this is snagging whatever it can it charges itself up. And then it has some kind of transmit capability. Yeah, but it does it without actually having a power source. That's the cool. Yeah.
And it's like it runs on the whole system runs on like, 3.5 micro watts. Yeah. And it can transmit the 30 was 30 ish feet.
Right. And the transmit that it does is really distorted and garbled. But it's intelligible. Yeah, I mean, you can actually decipher it by hearing it. And that's impressive. You know, eventually, I would, I would think that these kinds of things would be more feasible. Yeah,
I think it would have been cheaper to take to silo cups and a string. String it up and actually you might sound but it's got
to be electrical to be on Hakodate like that. You know it for it to be really like that. So it would have to be a wire instead of a string in between the cell. Oh, yeah. Yeah.
Okay. That's cool. Yeah. Next one is the evaluation boards for USB type C power delivery found on electronics weekly. So I've been looking for a long time the mythical unicorn of the basically a example of how to do 100 Watts, just real 20 volt five amp board Correct. over USB type C power delivery. And our own announced that they have a dev board that you can buy the evaluation board.
actually rip a full 100 watts. Yes. Nice.
They have a whole bunch of stuff that goes from 15 to 100. They got like 15 different evaluation boards for all their their junk, basically. USB type C junk. Yeah. The I think it's called Yeah, BM nine to a and then there's a bunch of other numbers that come after all the other different flavors. Yeah, all the flavors that comes in the one that you want to get is the BM 92821 M WV dash Evie k dash 001 Because that's the 100 Watt one
I you know right now I still don't know of a good application for that. Um, I know we talked about a soldering iron at one point in time
the fact that you can charge your your cell your laptop off the same port that you can read that out of Apple really likes that idea.
Huh, but Apple probably doesn't like the whole generic connector idea
now they use in their latest MacBooks have just one type C port
no lie. I don't know that. Yeah, I would I would totally have expected Apple to make some kind of proprietary crap can can
USB type C connector that's like slightly wider. Well, yeah.
USB type C Apple ca or something like that. And it cost you no no apple 50 bucks for just the connector. Apple C apple. See that's what it is. Yeah.
So I started digging in Yeah, cuz I found this and I'm like, I want to buy one. Availability now pricing. TBD. I'm like
prayer Yeah.
I just through my paperwork for
pricing, if you need to ask. Exactly. Yeah. Ah, so we need to make our own type See 101
I guess so. You can if you go like check availability on their site for all their distributors, Japan their Japanese distributors got like 12 in stock of all their stuff. So I might try to actually buy one.
What is it? What does it take? You know, to to actually ask for and get 100 Watts off overnight. We
talked about it. There's that project the USB PD buddy. Yeah. reparse should go look into that project and get that because you can buy it. Okay, I think he actually we built his stuff to
do we really there's
a plug for Makerfaire right there. So go buy some by his by his stuff. No, I'm gonna go until
he gets it done.
Going by one and see this his platform supports it.
An actual, like, honest to God 100 Watt
No, no Is it his is just like talking to the system. Okay, so you can request 100 watts. Gotcha. And so we had to make a
doesn t I have some kind of power management chip that does type C?
Yes. I was actually playing around with it, but I could never get up to 100 watts. Okay. So I, I want to try to order one from Japan, and we'll see how long it takes to show up. Okay. I can use the marketing budget right for that. No,
no. That's pretty great. No, no, you know. So, remember, we saw this a couple weeks ago? It was power filtering for PCI Express cards. Yes. Oh, we didn't talk wrong. We did not talk about but I think this is a good situation. Basically, there was some goofy audio whatever card they were PCI Express cards that were basically like the size of a video card, like a large video card. And they were just jam packed with like, just capacitor guy scrapers, cityscape worth of electrolytic capacitors. And they were just power filtering for the computer. And it was supposed to be some kind of like, improve your computer's sound. Okay, great, whatever. But for 100 watt power delivery. That would be awesome. Oh, yeah. If you could store up all those caps such that if someone asked for 100 Watts, it got it immediately. That would be really cool.
Or blow some stuff up?
Yeah, that would be the PDS the power delivery system. Yep. Yeah. Cool. And with that, now the Yeah, the RFO is done now. Yeah. So that was the macro fab engineering podcast. We were your host, Stephen Craig and Parker Dolman. Take it easy, guys. Later
PCB serial numbering? Parker and Stephen cover their thoughts on applying a unique identifier to PCBs in production for inventory and testing control.
This week's topics are: Porsche's Synthetic Gasoline, Record Chip Manufacturing Sales for the year 2022, and the Raspberry_Pi Social Media Firestorm.
What are the common Design Rule Check errors that Parker and Stephen see as Contract Manufacturers? Are these DRC errors the ones that you run into?