The US Mint Denver produces 30 million coins a day. Denes, the tooling department manager, discusses with us how production at this scale functions.
Stephen is on the hunt for the next step in his electrical engineering career and shares the shifts in the industry and what employers are looking for.
Relay manufactures hate this one simple trick that makes your “sealed” relays last longer! Except TE connectivity who has an note about this relay feature.
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 host, Parker,
Dolman and Steven Craig.
This is episode 247 47. So for I think, Episode 255. I'm going up to Colorado to hang out with Steven.
Episode 256 Is Christmas week. So it's the week right before Christmas.
Oh, that's gonna be interesting. So you actually looked at it then?
I actually just did like a minute ago. Because I'm like, hey, when does this actually happen?
So it's the week before Christmas?
Yeah, let me see here. That's December. The 17th? No, sorry, December the 15th.
So let's figure out if it's gonna work or not. Yeah. Yeah,
we got to work something out here. Yeah. But the eight bit episode that's coming up. Yeah, the
eighth episode, and then we'll have to wait quite a while for the 16 bit episode. Oh, good Lord. Don't think there is a electrical engineering podcast that has that many episodes yet?
I think the universe ends before that happens, right?
Oh, no, what? No, we got how many more episodes? So
the heat death of the universe? It's more than 16 bits worth of episodes, that's for sure.
Okay, so today I want to talk about PCB stackup. But not like, all encompassing. This is just kind of a question that came up in our engineering group at macro fab. And it's kind of like one of these. Something I want to ask you as well, many, because you also do contract manufacturing for printed circuit boards. And so my question is, what do you do when a customer brings you a design and doesn't have a stack up? So first of all, what a PCB stackup is, is basically all the thicknesses and all the layers of what makes up a PCB. So you have solder mask, you have copper, you have ecin, which is like, you know, in egg, what are your surface finishes? You also have the Peg, Yeah, prepreg, and core, which is the material that you're the fiberglass material. Most time it's called fr, four. And there's different ratings of Fr fours. There's also like Rogers, and bunch of other stuff that's mainly used for higher frequencies and that kind of stuff. But so, if a customer does not have a stackup, what do you do?
Okay, so I think one of the main differences between my contract manufacturing and your contract manufacturing, is that mine is very specific. We manufacture one style of electronics, basically. Yeah. So we we actually very rarely get questions about the, about the actual stackup. We've, I can think of to customers off the top of my head who defined those. And those are pretty much the only two customers who have provided fab drawings that show the stack up, and the stack up that they provided are just standard stack up. Yeah, there's standard setup. So so the it is, so in the in the synthesizer world, and even the Guitar World, it's it's way way behind in terms of going to high speed, and I guess, high technology and basic boards that require impedance control and things like that, because most of our circuitry is in reference to audio frequencies, unless somebody is doing some DSP stuff, which is, which is a bit faster, but even that is not as intense. So that being said,
for the most part, we, our customers just have an assumption that the PCB that they're getting is a standard stackup. And we just go with that unless they require something different. So in terms of our I feel like we're a little bit unique, because our customers come to us knowing that we're specific with that industry, but if say, so a customer is going to macro fab, you guys are generic, you guys will do anything. So there's a little bit more of impetus on the customer to provide their own information on that.
Yeah, this basically happened was the example would be a customer came to us without a stack up and we asked for a stack up and the customer was wondering why they had to provide a stack up which I thought was very interesting, because this is the first time in years this has ever happened to me. And and the reason why I gave with why we define it all the time is This our sales team is we want to make sure when that customer comes back and orders again, they get the same product again. Because if you if you ship off a board to get fabricated somewhere a PCB fabricator and then say you ship off that board again to a different fabricator, they might use a different stack up if you don't specify it.
Well, what No, no, doesn't matter if I have a default stack up, or I know you guys used to
know we had the vault socket, just this board did not fall into the stock stack ups, because they had a weird combination of of copper waits for inner and outer layers.
Oh, okay. Okay, so that makes it nice. So when you first said that the customer is like, Why do I have to provide a stack of I in my head? I was thinking the same thing was like, Well, why do they have to if you guys have a default, but if they've deviated from the standard, then they should provide?
Yeah, they were deviate, they deviated on the copper weights. And so we were said, we asked, you know, do you have a stackup? And they were asking why they needed one? Because they're like, well, we just want two ounce copper everywhere. So they wanted to answer on internal layers to internal layers. Yes. Oh, wow. Okay, that, you know, that's something important to note, like, if you go with thicker copper on your internal layers, that does change the thickness of the copper on the internal layers. And that changes everything really. So
that's really interesting. I've only once run into a customer who needed thicker internal layers. And when I say needed, they just wanted it. They didn't
need it. I've seen I've seen like one ounce, half ounce. I think our standards are one ounce, one ounce.
Yeah, I believe that's right. I that was one of the things I did at macro fab was was right up a bit of the standards. And I think it was one one at the Fed.
Yeah. I took I actually took your stuff probably modified or two years ago, and I made a really fancy looking PDF out of it. Oh, nice. Yeah. So I'm gonna look that up real quick. Make sure I'm not lying about one ounce, one ounce.
I think it was one ounce, one ounce, because a lot of places do have one ounce, half ounce. Yeah, it is
one ounce, one ounce. And I think we offer two ounce and half ounce. It's just we don't have standard stack up. Because usually when someone picks those, that generally, I'll say 99% of time they have their own stack up they want to use because they're picking it for a reason. It was just this one incident. One case, I guess that a customer asked for asked for a really weird copper weight stack up and didn't provide a stack up and now we're like, Well, can you let us know? Like, what, how we should build it and know like, why?
Well, you know, it's actually it's funny, because I think this is where macro fab does things? Or is changing the game a little bit. And, you know, in some ways for the better maybe in some ways, not? I'm not exactly sure. But traditionally, whenever you would get a quote for a PCB, you would send out whatever the files were the film files for, right. But you'd also send out a drawing that was your fabrication, drawing, and your fabrication drawing, had all the pertinent information. So all those buttons that you click when you just order something on, on, you know, whatever webpage, like green solder mask and white silkscreen and things like that, those were just written IPC class two, right, right. Those are written notes on a fabrication drawing. And that's the classic way. And almost every fabrication drawing had had a stack up on it. And that's how you mitigated or got rid of the problem of if you ever ordered this again. Well, it's right on the fabrication drawing, you're sending it again, you always get the same thing.
So this is what I've noticed on those fabrication drawings from customers, and I'm about to call out like a lot of people in this world is, especially engineering firms about to call out is, like, half the time you'll look at it and be like, why is that? A? Why do they call that out? And so you ask them about it in that oh, yeah, we just put on all our drawings. Yeah. It's like, it doesn't matter. Don't worry about it. It's just like, really,
that that means that the like, I don't want to get too much of a stickler for it. But if something is wrong on a drawing, then you change the drawing. If something is wrong, and someone says don't worry about it, that instantaneously means everything else on the drawing is worthless.
Yeah, exactly. I you will not imagine how many customers I've, I've worked with that had that had that problem. So I think the way so in McAfee, basically, we were saying changing the game. It's more about taking those drawings and stuff and then you basically were basically making the point With a drawing is you have to go read the drawing, and to figure out what's going on. Whereas the macro map platform allows you to set all those options so you can automate those options.
Right? I we've said this before, the mech fed platform is an automation of a fabrication drawing.
Yeah. And so that's, that's what we're, you're, you're translating all this old stuff into a new style format. And so the way I think about it is standardizing how you define a PCB assembly. That's kind of like how I view the Mac fan platform, in terms of from an engineering perspective. But yeah, it's amazing just how many fabrication drawings you go through and like your why's this? Like, like, or sometimes like, oh, yeah, solder mask should be green, and then they send it in their email. Oh, we want it blue. Now. I'm like chains are drawing then.
Yeah, exactly. You know, it's funny, I was having an argument with someone about the term living document. Just the other day, and I was I was kind of, I was arguing about things in an engineering perspective, where I believe that like, take fabrication for drawing, for example, I believe, that constitutes a living drawing or living document, but a living document. Has is has snapshots. So in other words, say you have rev a of a drawing, and something needs to change, you make the changes, and you rev two, Rev. B. Now, Rev. B is alive. Rev. B is a different drawing than Rev. A, it's not. It's not the same drawing, like the person I was arguing with was saying, like, No, it's the same thing. It's just that Rev. A is alive. And it's now being like, no, because they're actually different, in my opinion. And I think you get into trouble. If you start thinking with with that other mentality, like, if you make a revision, rev A is dead, and it gets archived. And you can go look at it if you need to, in case something is wrong with Rev. B. But Rev. B is now the main document. So like, in my opinion, a living document is one that can be changed, but it is also fixed. You have to go through the process to change it. And as soon as you change it, everything else gets archived, you know?
Yeah, the way the maquette platform works with that now, because there is like revision history and stuff like that, is it works. Because it does do snapshots. That's how like revisions and versions and stuff work like that, but it does it by recording the changes made to the original. Okay, so it's kind of like, it's actually a lot like how GitHub would work, where GitHub, you when you initially upload all the original files. And then it's only after that it's recording just the changes, for the most part. If it detects like massive changes, it will just do a whole, like rewrite of the other file, but it's mainly recording the changes to the files, which ends up being less storage in the end.
That yeah, that that makes sense. I mean, I from from a software perspective, that makes sense. I think from a, from a hard engineering perspective, I would want individual files, one is A and one is B.
Yeah, the Gerber files are treated that way. Because it's doing a diff line by line of a Gerber file means nothing to inhuman.
Oh, right. Right. Yeah, exactly.
But like specifications and stuff like that, oh, especially if she got turned from Green solder mask, the blue solder mask now on this date? And the you know, what, who did
it? Actually so so I think I think what you're capturing is revision history. And the snapshot can be built instantaneously. Whereas with things that are more complex, it's just easier to keep a hard copy of whatever
it is. Yeah. Yep.
All from PCB stackup Oh,
yeah. Well, not even like going deep into PCB stackup. Just like, wondering, like, defining it, like I am, I'm the kind of person that defines a PCB stack up for like, any Joe Schmo two layer board. Because I just want to make sure if I ordered that way, before I it always comes back that way. Right. Right. And I think that's really important for your next topic, because it deals with FCC CE compliance stuff, but that's that's the important thing is like even if you do a Joe Schmo two layer board, and you get FCC CE compliance on it. And they you build it somewhere else, and they change the stack up slightly. That's gonna change the performance of your board. Rather than not that it's banking on like impedance control, whatnot, but it gets sold Cause uh, you know, it could just be right on the edge and now it's radiating too much.
Yeah, sure example. Sure. Yeah, it's a lot of those things you don't know until you test it, right? Yeah. But But yeah, you would want to the best of your capability, you would want to be able to define and control your stackup. And, you know, that's another thing that like,
in like I was saying earlier with macro fab, like kind of redefining the game on that. Because of places like macro fab. A lot of people don't think that they have to define their stack up because they just click Oh, it's for layer. Good. I'm done. Right. Whereas previously, you were required to provide a fab drawing like places wouldn't give you a quote without a bad drawing, because an email wasn't enough.
Yeah, it went, it went from that being the requirements. And then basically, we went into the wild west of PCB assembly of not assembly, but PCB fabrication where you can just send your drawings like let's say, OSHPark, for the example, as you send your board was going through my head. Yeah, it was like you'd send your files out and you just get a two layer or four layer board back yet like it didn't You didn't care in quotes about the stackup, because you're getting cheap boards. And so Mac fab is trying to I guess, crawl crawl back to the let's specify stuff. side of it. Again, not not the nakash OSHPark. Because we use OSHPark. That's actually the first couple of years. They were our fabricator. I've certainly
used my fair share of OSHPark boards. Yeah, I love OSHPark boards. Frankly, I don't know off the top of their head. I don't know what their stack up looks like. But I'm sure they just use the standard.
Right. As far as just some standard stucco. Yeah, something like purple on it. Now, here's
the hard part is being able to determine when do you need to define a specific, a special stackup? You know, I mean, obviously, our definition of like, you change copper thickness on the internal layer as well. Now your stack up is different. Right. But like, when does it become important for your circuit? That's a lot harder to determine.
Yeah, we should, especially since I don't really have a lot of experience with with that kind of stuff. We should probably get someone on podcast,
that would be fun.
Probably confusing more than fun, right? Yeah.
All right. So back to FCC compliance with Brexit dunk dunk.
So this is this is a fun one. So actually, I keep
throwing them that F word around. No. So talking about
I, it's kind of funny. I was, I was joking with my boss today, because he sent me he sent me an email. And it's from ul ul is holding a webinar coming up here soon. And he's like, Hey, I thought you might enjoy this. And of course, like, I'm that kind of nerd was like, oh, yeah, I actually want to see this. This is kind of cool. So I signed up for the webinar. So it's all about Brexit, and the CE mark. So Brexit being the UK, leaving the European Union and the CE mark being the mark of the devil from the European Union.
So So Brexit is not going to be beholden to all of the requirements of the CE mark. So as with evidence, it's gonna
be the wild west of selling electronics in the UK now.
Well, as with everything Brexit everyone is like, what do we do? Like, if you know anything about Brexit, or followed it at all, like, it's just, you know, depending on what side of the aisle you're on, you're either like, super happy about it, and you're or you're just like, this is chaos.
It can be both
that here's, here's the thing about it. The the the the the UK, leaving the CE mark means that they are in the wild west. So they have a couple of choices. Do they maintain the CE mark? Do they require that? Do they require nothing? Or do they make their own mark, and I'll let you to kind of guess which one they went with?
It's the UK and they never left the British pound? Yeah, yeah. They never ever went to euros, right? So they're probably going to revert to whatever they used to have. If they didn't have anything, they're going to make something new.
They're making something new. So we got to have
one standard for everything. And now we're going to have four No, we're gonna have four Yeah, we'll have four standards because we want to have CE FCC. This UK thing and we have a Japanese a Japanese one too.
Oh, there's way more than that. So okay, so like talking about the ones that really matter wouldn't get this here are marks to keep in mind right now. Okay, FCC, UL, CSA. C. And in. Now this new one that is called UK ca.
Well, I don't think you can sell something without a UL Mark. work anywhere?
No and Yeah, well, or ETL? Yes, yes. Well, ETL is also kind of another because it's sort of like a amalgamation of a bunch of
Yeah, I view you Well, an ETL is kind of like safety certification stuff.
Well, those Yeah, that's what they do.
Yeah. Yeah. Well, I'm talking about compared to like, FCC or CE mark. What
right? Well, I mean, UL is more concerned with safety, in general, and C is a little bit more about the polluting the environment with electronic crap.
Yep. an FCC just cares about what you're radiating electronically, or electromagnetically.
And CSA kinda does a little bit of both. So so the whole The whole point is, when I was saying, like, these are the marks to keep in mind, like, if you're developing a product, like you got to know about all of these things and how they apply to the markets that you might want to get into. Because, say you want to, let's just pretend you wanted to carry something at a big box store in the US, well, you will need the Well Mark, because most big box stores require the UL Mark, you're guaranteed to need the FCC mark. But if you're not selling in the EU, you don't need the CE mark, because that doesn't apply on this side of the pond. But if you wanted to expand into the UK, the CE mark isn't going to be your friend, you're gonna need this new UK see a mark. But here, let's add some complexity to make it more fun. That doesn't necessarily apply to Northern Ireland. Which they are reverting to the CE mark. So it's all just a mess now, and you know, just to make it easier, like why don't we just have another body that you have to go do?
I do like the UK CIA's logo? It? Yeah, it's got that like, it's got that like cyber punky font. Just a little bit. Yeah, yeah. Yeah,
I like that a lot. It's way better than C will. And here's the thing, we don't have a lot of time. So if you weren't aware of this, I'm sorry. This is kind of sucks. But so this all goes into effect. January 1 2021, is when the UK see a mark is required on your product. Three months away, there is a grace period where the CE mark will still okay, basically what it is, is the UK ca mark, it will be accepted January 1 2021. So that's the first date that it it actually goes into effect. You have one full year where the CE mark still applies to things sold in the UK, when and January 1 2022 is when the CE mark gets phased out and the UK ca mark which at the beginning from from everything that I've read the UK ca Mark seems pretty one to one with the CE mark. And I think that's done.
They just copy pasted the document. Well, I
yeah, I think they did that on purpose to make sure that the trends, the translation and the you know, getting into things is smoother, as opposed to like, here's all these brand new things that you have to do that are all different. I had all
your PCBs must have a picture of the Queen on it.
That's coming in a couple years. But I have a I have basically www.gov.uk There's an article about like, the entire UK ca mark, not uk.gov It's gov.uk.uk Yeah, okay, so this this article here if you don't know much about the UK ca Mark, this is a good place to start reading about it with some of the requirements some of the unique aspects about it. What they it's very much like the overview bullet point, the front page of what seems good seems like it's going to be a massive document you know, because it's like re rewriting CEE you know, so it's gonna get pretty nasty. They don't go and do a huge amount of details in this page at least
yet as I was looking at it was like it's it's anything that would require a UK see a marking, it's like what is that? Oh, there's not a list yet.
Now Yeah. Everything it's this is very much Brexit you know, everything is fresh, everything is new, like how are you going to handle it? So what this is meaning is get your stuff in and you start learning about this as much as possible. So what's nice is if your product can get the CE mark, before January 1 2021, you can sell it in the UK. There was there some chat about if you're if you get the CE mark afterwards. that you have some troubles because they're going to start requiring the UK ca mark. Also after a, I can't remember what the date was they haven't written on that page, which will provide the link in the show notes. The UK ca mark will be required to be a permanent mark. So you either have to mold it into your enclosure somehow or you have to have it in an ink that can't be worn off or lasered in or something. And I'll say no more stickers, no stickers, yet they it but that's not an immediate thing. I think that's a few years off.
I wanted the silk screen counts. I think it does. Because it's permanently in quotes attached to a PCB.
Yeah, you might have to prove that it's like an epoxy silkscreen ink or something. Yeah. So
something that, like, I'll call wooden wouldn't remove.
So one of the things I was reading, that's that's important, especially for someone who maybe hasn't gone through regulatory approvals before they even talked about it in this article is that documentation is key. So you have to prove documentation behind your manufacturing where you got your stuff, how your thing gets manufactured. Like, if they pull out a product, they have to see your manufacturing documentation behind that.
So Wow. So it's even going farther. CTE doesn't go that far. Yeah, yeah, they.
So at the same time, it's important to note, I should have said this at the beginning part, neither Parker and I are experts at
all. Ever heard of this? Yeah. So
don't take our word for any of this. Go and learn about this. yourself. Go research it. But yeah, this is sort of the first that I've been learning about it.
Oh, no. Well, are you talking about the declaration of conformity?
Well, you the declaration of conformity is something that you have to write Yeah. Yeah. But that's that's similar to see ease. Yes. But But I think in the paragraph above that, they talked about your manufacturing documentation. Gotcha. Yeah, I'm reading that. Yeah. So I don't know the extent of that. Because I don't know if it's like ISO required. I mean, two hours,
it just like, the product shows up at your door, and you have to start documentation because it's talking about you as the manufacturer, right. Not your CM.
Oh, no, I don't think I don't think it's you. I don't believe it's your CM. That would be awful.
Yeah, you have to track your printed circuit board all the way back to like the mines that the copper and the fiberglass
came from. Yeah, no, it's not that intense. That's actually that's a good question.
What is like, where
does it stop?
Yeah, we're gonna stop. I, in the beginning, there was the big bang. That's like the first line of every single document.
What was that? What was that Carl Sagan quote? Like to bake an apple pie? You have to create the universe or something like that?
Something like that? Yeah. That's good.
So the UK ca mark, it does look really cool. It looks cooler than the CE mark. That's for sure.
So go go check it out. And I'm looking forward to this webinar. If you go to you wills website, somewhere has this webinar and we got an invitation through an a, an email, but I'm sure you could probably find that out. It's scheduled for this month sometime. So I'm looking forward to that to just be like, Oh my God, what do we have to know about now?
In the in the question to answer section, you should ask the person who's doing the webinar if they want to come on the podcast.
Oh, that would be cool. I might have to do that. You got to do that. Okay.
All right. Um, next topic. So this was something Steve and I were talking about before the podcast. And so when you ferment beer, or wort, you're turning wort into beer, where the fermentation process, you have to keep it at the right temperature. I think we've talked about this a couple of different times. But yeah, you need to keep it cool. Because the yeast have an exothermic reaction as they're eating all the food because they know that metabolism, right so that, you know the sugar can now call and waste is heat out of that process. So you have to keep it cool to keep them from overheating and making off flavors. So there's a couple different ways you can do that. Most home brewers have a ice chest or a freezer, that they put them in there that they put your your fermentation vessel in there, and then they you set the temperature with a digital controller. Good to go. Works great. That's how Steve and I brew. What combination 20 years probably brewing. Yeah, probably actually that's somewhere around there.
And actually one of the one of the aspects of this is not necessarily the absolute temperature Like, that does matter, but what matters more is maintaining the temperature than, than the actual absolute value.
Yeah, correct. So there's the problem with doing it that way, with basically just having a big chamber that you put your fermentation vessel into, is, you can only technically, do one beer at a time ferment one beer at a time, because if you put another fermentation vessel in there, usually it's not at the same time. So it's like, down the road. And you might be at a different phase of fermentation and one vessel versus the other. And they require different temperatures if they're different kinds of beers. So you pretty much have to do like the averaging game of like, well, I need to be kinda in the middle here, and you don't want you know, heaters to be fighting each other inside the inside the freezer. And so what what professional brews have is a what's called a glycol system, and they have a water jacket that goes around their fermentation vessel that cools the entire vessel down individually. So each chamber has their own fermentation, fermentation, cooling jacket, control,
right? So say you're wanting to brew on the same day, a logger and an ale. The difference between a logger and an ale most of the time loggers are from an fermented, fermented, fermented at temperature ranges of like 45 to 55 degrees Fahrenheit, and then ales are like 60 to 70, somewhere in that range. If you wanted to brew both those on the same day, you'd have to be able to maintain that individually.
Yeah. And so a couple years ago, probably about five years ago, there was this big movement and people building DIY glycol chillers for homebrew and it's kind of a slightly different way, instead of doing a jacket around the fermentation vessel, they do an immersion cooler, which is basically a big stainless coil that goes into the the fermentation bucket. And that seems to work just fine. And there's been a couple products that's come out that that are like DIY style. glycol chiller is basically like the like mini fridges that, you know, you have pumps and stuff in it. So Steve and I were looking at these and the problem with making a DIY one is it looks like a DIY one. Like it just doesn't look good. They look off because Yeah, cuz you take like $100 window AC unit, and you hack it apart. And then you take the the evaporator and like bend it into a cooler, and then you flood the cooler with glycol with a submersible pump. And just looks like eggs, and then expanding foam The Seagull. Oh my god, this looks bad. I'm sorry for if anyone's actually made a really good looking one. Let me know. Because I might want to copy your design because everything else I find is just like, yeah, that looks. It works it apparently they work great. It's just man, they just look ugly. And especially after spending, you know, seven I spending so much time actually making our rigs like look nice. It'd be a shame to have that sitting next to it. Right?
Yeah, you don't want this like even though there's there's fun things about like seeing the guts of whatever project you're working on. Like what we've described in the past with like thunder homestyle electronics, you don't want a thunder own style, glycol, chiller, it just Yeah, it doesn't instill confidence in the whole system. Especially where
like, most time when people like oh, like you offer someone a homebrew. And they're like, No, right? Because, you know, there's this stigma about homebrew being, you know, bad.
Yeah. And then they look at it, and they're like, Oh, so you're pumping weird chemicals through a bunch of tubes on the floor. And then I just see it, like, go into this beer and come out and like, yeah, they're gonna they're gonna look at that and just politely say, Well, maybe not even politely they're just gonna be like, no, no.
So so we started looking at that the I'm doing air quotes prosumer models, it's like 800 bucks. You can get like a black mini fridge that looks like a it's a glycol cooler. So we were like, Well, I wonder that's a lot of that's a lot of Jeep parts is what I was thinking was thinking. And so we started looking at different ways you could chill like water. And so Stephen found this thing called the chiller daddy, which might have the worst name of all time.
Oh, come on. It's so great.
And it's only it's I say only it's $350. So it's cheaper. Yeah. Now it doesn't have a pump because basically this unit is designed to go under your under your sink and cool the water, which I thought was really weird because the like, I guess if you don't have a fridge that has cold water coming out of it, but I think every single fridge has that now. I don't know. And I know you can buy a fridge for under $350. So anyways, I thought it was really weird that this product exists. But you can take one of those. And then like, Steven has like a $27 pump that you can buy,
which I have that pump from for brewing. And it the day I bought it, I tried it out, or I've used it on multiple brews now. And it kind of upset me because I bought this $27 pump off of Amazon. And it's as good as my, like $250 pump. It's like it's like a fourth the size and it has the same flow rate. And it just, it's great.
Yeah, don't don't make because I have three of those expensive pumps. Now, so take one of those. And then basically you just cycled the glycol into the chiller daddy. And then we had to figure out a way to divert the glycol into the immersion chillers and all the vessels. And so I was actually thinking like sprinkler. Sprinkler valves. Yeah, yeah, that would probably work. Did and then couldn't have that on a on a temperature. Temperature control.
Yeah. So basically, you have X number of fermenters, that each one has a coil in the lid that goes down into the actual liquid. And then once it reaches a certain temperature, you will just click on a solenoid valve, pump some glycol into it until that temperature is where it needs to be and then just turn that valve off. Yeah, yeah. Should work. Yeah, you know, something that was just going through my mind. In, in brewing, a lot of a lot of the ingredients that you hang on to, you need to be cold. Or if you harvest your yeast or anything like that, you need to keep them in a cold environment. Maybe there's a much easier way of doing this, could you not just buy a really cheapo mini fridge that has a freezer, put a tank of glycol in there and then plumb some tubes into the side of it. And then it doubles as a little storage thing for all of your hops and your yeast. And it's got a little tank of glycol up at the top. And you could probably buy that for like 100 bucks at Walmart.
Or cheaper. Yeah, yeah, that's actually a really good idea. And, and it would look good too. Yeah, cool. Just look like a mini fridge with two hoses coming out, which wouldn't look bad. Well, you
could, you could do some nice fittings on the side of the fridge and then it would look purposeful. Hell you do it off the back, and then you wouldn't even see it.
Yeah, and then the Yeah, cuz generally, yeah, you can, you can put the glycol in the freezer section. So it's at like, you know, 30 degrees. Exactly. 50 or whatever the fridge decides that.
That's a far wonder. Yeah,
I wonder what I guess you had to experiment with it. But what your beats you output would be or input I guess, I would say I would, I'm wonder like, like, I wonder what the what i How much joules does yeast output?
Yeah. Oh, so okay. I do know. i This doesn't answer that. But gosh, I was talking with Brock. I can't remember his last name the guy from St. Arnold's in, in Houston. He's the owner of St. Arnold's. And he was talking about homebrewing the temperature inside the warm safe, you just put your five gallon bucket in a fridge and you said it for 60 degrees. It can be upwards of five degrees higher inside the water inside the actual Wort than the out the external temperature. So whatever. You know, that doesn't necessarily answer how many joules of energy it is. But that's a lot of temperature rise for for just a bunch of bugs eat and crap. You know,
you know? Whatever freezer. Well, it's a my chest freezer can easily keep up with like, four five gallon buckets. That's probably fine. It appears to be just fine. I bet
you you could get away with the whole project for like 200 bucks. Yeah, let's try it. Yeah, because I will totally do that.
This is a good idea. Yeah, because they're gonna go to Amazon shopping after this podcast. Just buy
a little chest freezer. Yeah, it's even cheaper. No, no
fridge, well fridge with the freezer, because that way you can keep your ingredients in In the bottom Yeah, that's such a good idea. That's way better than the chiller daddy. And it also doesn't have the name chiller daddy on it.
Come on, that's such a great name.
The website for that product is a little creepy, is it? Yeah, it's well, it's an old old website, but just has like stock photos, but they're probably not stock photos are probably pictures of whoever's the person who runs it family,
whoever's the daddy. Yeah, whoever.
Whoever is the data, yeah. Okay, last topic for the day. PCB badges. So what I've noticed on the electronics subreddit is that there's like the war of badges, like people like one upping their badges, like, oh, one person used like an O for two instead of, oh, 805 on this package, and then like it's escalating from there, like through the roof. It's like wanting pictures like, oh, I put a dip 60 on the src 16 package. And so my question, Stephen is, what is your worst Bob's that you ever had to do? And what is the most proud bodge you ever had to do?
Oh, okay. I guess this,
those are two. The two examples, I gave her, like different packages on different footprints. But this could be like wire barges or anything like that.
Gosh, because I've had my fair share of budgets. That's that's for sure.
If I remember the first revision of the that, that synthesizer he made right before you went to Colorado? That was baje heaven. Oh, yeah. baje hell, I guess Baulch Hell
yeah. I mean, it was it had 600 components 600 Plus components on it had like 12 sub circuits, all untested. Like it needed some logic. So yeah, two was pretty good. So yeah, yeah, it all worked out.
Gosh, okay. So like, the thing is, it's funny because I'm kind of known for bodging at work, but like, I'm not known for doing shitty budgets at work. I'm known for being like capable of lodging at work. Like, every one of my prototypes. You got a bunch something you know, like, it's not a prototype. All your badges are proud budget. I'm pretty happy with my budget. Here's one that was that was fun. This was back in the macro fab days, I was making a test jig. And I don't remember exactly what happened. Somehow something got screwed up where some layers got connected internally on the boards and I don't remember how it happened. I took out a drill bit and I was drilling out vias to disconnect things and then I ran some bodge wires on top of the board. And that worked out that worked out pretty fine. And I remember being like real careful with the with the via holes, because I didn't want it to like have chewed up edges or anything like that. So unless you knew what what you were looking for, you wouldn't know where those Oh, you know, I think if I remember, right, they were I flipped like SDA and SCL on it. Oh, I do see lines. I believe that's what I did. And I had to I had to flip those around. So I had to drill out there was luckily two vias right next to my processor, so I drilled those out and then flipped them around.
So my most proud baje I guess, is we had a customer built, we built like five prototypes of this board, so everything was fine and then we built like 250 but some communication they updated the version in between and I'm like, we should run a prototype and like now it's run it ran it they they moved one of the parts. Okay, well actually though, they mirrored one of the parts that put on the backside and this was KY CAD. When you do that in KY CAD, you have to rerun your polygon pores. Oh, you export Yeah. And so it can drills stuff went straight through all the planes. And so yeah, did that I basically drilled out all the all those holes, because they only actually needed the center one for some reason. For some reason that it was like they only need the center one to actually be connected to something. And yeah, drill them all in but the second one was like the same so it didn't actually need to be drilled out. It didn't connect anything besides like a top trace. But yeah, it was like 200 for the boards times 10 drills. Oh, that sucks. But hey, got them all working.
And you know, I had a yield like this one I had out of board that I needed to send. So basically what I did was I had some transistors circuits on the backside of the board, where I kind of foolishly just put some transistors in parallel, got the board made. And then I was like, Oh crap, basically, I made logic where either transistor could make a sub circuit function when I needed them to be independent, but I put them in parallel like an idiot. And so I had to cut some traces to make these transistors independent. But now that they're independent, I needed to run a new line from my processor to one of these transistors, which luckily, I had plenty of extra GPIO. But my, these transistors are halfway up my board. And my processor is at the very top of the board on the other side. And there was not a good way to get around. And I needed to barge a wire all the way through. So I got really, really thin like 28 gauge wire. And I made super nice like turns and 90s and things and I snaked it through one of the holes of the of the my tag connect programming header, and then all the way around. And it was so good that you could still connect it a tag connected to it and program it so that was a fun bunch.
So the one that I think is kind of interesting bodge wires is way back in the day. The when I was doing the pin hack pinball systems is before the penetrator. We were on Revision three, which was like so prototyping and figuring out like how to make a pinball machine work. Sure. And so we had a board that would we barged into rev four or the board. And so read for was production was was the production revision for America's most haunted pinball machine, which is the first pinball machine spooky pinball made. And I think we built like 100 of those. But but the first couple were built from bodged rev three boards, because we wanted to like actually get some pinball machines in the wild to actually like test stuff. And there is one that still runs because all the other ones got like recalled, not recalled. It's the wrong word for all of them got pulled back in like we like gutted and then put all new stuff into it. But one is like at a pizza parlor like right next door to spooky pinball. And that one is running like a rev three board still. And it's still running like the route because the thing is we can't upgrade the code either. Because like we did do some like pin changes. So the code is not compatible between red four red three. So so running like the original old code. But it's still running today. It's that's got been like seven years ago now. Wow. Six years ago, six years ago. That's awesome. And it's running badges. Like because we would we ran all the barges and like hot glued them down the wires down so they wouldn't vibrate or anything. And then that things so apparently it's like, still trucking today. Wow. So that's the longest running bodger ever had I guess.
I don't think I have any barges in the wild. All of my boxes, our product. Yeah, most
of them. They're on your desk or whatever. But like this one's actually, I think there's a couple the red three boxes are like, spooky promos got like a glass cabinet. They put their history in. And so they got all those boards in there, except there's still that one in the wild. That's it. And they just don't want change it. Because that was like the first that was the first machine they ever delivered in quotes. Nice, too. So it's like, they put it in there just so people could play it. Test it out. Very cool. Wonder how old that code is? Like, how different that code is from the would you even be able to tell? I think yes. I know there's been like, I think Ben did like 30 or 40 revisions of code. And like even when that rev three was that like even the artwork is different on that game. So it's way different playfield I think is pretty close to production but like all the artwork on the outside and stuff is way different. Like I think we were like it's on the outside of the cabinet is like Photoshop people and stuff. It's not like drawn or anything. It's really It's interesting. So I guess we'll wrap up so before we leave we want to see pictures of your badges in the Slack channel.
Oh yeah. Give us your worst or best. I think I have some, some bodies all that I have. I mean not think I've got bodies all around me right now. But I think I have some fun blotches that are I'll take some pictures and show you guys. I actually I, okay, got a fun budge. So designed a switch mode power supply the other day that functions properly like there's nothing wrong with it. And it's just a switch mode that takes 12 volts and drops it to 3.3 to power my digital stuff. But there is one setting on one sub circuit at a different part of the board that if you put it into that setting, then the switch mode supply messes with its operation and oscillates in a weird clipping portion of that circuit, the circuit gets flat, and in that flat section, it oscillates with the switch mode power supply. So I just took the inductor on my switch that switch mode power supply, and I kind of stood it up and soldered another one in series and I just doubled the inductance value. All gone, it just went away. So I just did a botch the other day and I had I had a customer or a client, I should say he has some of these boards on his end too. And he tried that also works on all of his boards. So just needed a slightly higher inductance value to get the duty cycle peaks down such that it didn't interfere with this. And it's purely a board layout situation that the current spikes that were due to the lower value of inductance screwed up another part of the circuit. So there's a nice, fun little badge.
How long did that take to figure
out like 10 minutes?
Oh, did you see what the frequency of the ripple was. And that was what your switching frequency was?
Actually what I did was I had the board on and I set it up in that situation where the ripple was there. And then I just started touching my finger all around the switch mode supply. And I could control the frequency. So it's like okay, well, it pretty much has to be that.
So how do you figure out that you needed more inductance.
This particular switch mode controller, the lower value of inductance is more suitable for higher loads. And this is a really, really light load. So basically what's happening is the duty cycle is working very, very lightly. It basically it turns on him and just slams a huge peak into so on the return current path in the ground plane is is lifting this other circuit, every one of those pulses. So by increasing the inductance, the duty cycle gets larger, so the current pulses are smaller, and this other circuit doesn't get affected by that. And it's funny because all the other the entire rest of the other circuit, it's dead quiet has like this SNPs doesn't do anything to anything else on the circuit. And it's a big six layer board and stuff. I wouldn't want to reroute things when I can just change the value of the inductance. And I was I was a little low on my inductance anyway, I could stand to go a little bit higher. So and that's
just going to drop your efficiency a tad, but
not actually it'll probably increase my efficiency attack because I was I was on the very low side. Oh, very low end cycle. And this, this should bring it closer to 50% which is nicer with this particular controller. Gotcha. Gotcha. It's a very light load on something that's like it's designed to hold to carry three amps of load and I have it doing like 40 milliamps, you know. Yeah, that would do it. Yeah.
Okay, let's wrap this up.
So that was the macro fab engineering podcast. We were your host, Steven Gregg
and Parker Dolman Take it easy. Later, everyone