Little Shop of Transformer Horror

Get all your turnkey PCB assembly services from Mac fab with new lower prices by an average of 44%. On high volume production, Mac fab has reduced prices on PCBs for production orders. This means that you will see price reductions for orders over 100 units, and the most significant savings for orders over 1000 units. Generally, we observe a PCB price reduction of 20% or for lower volume orders, and as much as 80% for high volume orders. Do you need a quote for high volume PCB assembly, send your quote package to sales@macro.com to get started today.

Welcome to the macro fab engineering podcast. I'm your guest, Josh Rocher.

And we are your hosts Parker, Dolman.

And Steven Craig,

this is episode 178. So before we get started, we have an update to the macro engineering podcast Useless Machine contest sponsored by Mouser electronics. So we've been talking about this for a while. It's basically a contest where you as our listeners, or anyone really will design a useless machine, upload it to hackaday.io. And our guest judges will pick the best one or the second best one or the third best one. And then we'll also have the Mac fat podcast favorite, which is one that Steven I pick a project. And those all have variety of prizes and dollar amounts associated to those prizes. So go check out our blog, it's a macro app.com/blog has more information about the contest and Enter. And we actually had our first entry by German. And it is, it's, it's pretty funny.

It's okay, so it's pretty great. It's technically not a complete entry. It's more of an idea that funnily enough in the entry is like, if people like this idea that I'll make it an actual entry. And, by the way, COVID Yes, you should. And this is our first entry and way to go for the very first one because Parker and I read it were like, Oh, this is great. The competition is already stiff. And there's one so far, which is great. So this first project is called the badge, killer shitty add on. So the description let me scroll down here and read it. What's more useless than a shitty add on a shitty add on that kills your badge, similar to the USB killer, but in the shitty add on form factor? Which park and I we had an episode gosh, probably 50 or 100 episodes ago where we talked about some cheapo Chinese USB killers that just shot high voltage down. Yeah, I watched

a video about that today from electro boom. Yeah, yeah. And

we've been talking about shitty add ons for for the DEF CON badges for a long time. So what about a shitty add on that? turns your hashtag badge life into hashtag brick life.

So Durbin, you should totally make this thing.

Yeah. Yeah. You already have the Parker and Steven approval on this one. Cool. So we have, we have Josh Roser. Again, as our guest on the podcast. He is now one of our more reoccurring guests, but also has some really cool stuff to talk about. Yeah,

you guys must be scraping the bottom of the barrel to have me on again. Don't tell everyone. Hey, as long as the check cashes, right?

Sure.

Yeah, yeah. Dollar Bill across table. George Washington, $1 bills. What do you think of George Washington?

Strange George Washington that's yellow colored. And he's wearing a top hat and has an eyepiece with monopoly? What the hell?

What? So the last time we had you on the podcast, you were winding an output transformer for a tube amp, which I think the day after we had that podcast that actually arrived at my house, and I installed it in one of my amps. And it is still actually, well, yeah, your house is not on fire. Yeah, my house is not on fire. It's still actually potentially still in the frame that these guys are watching on the video here. So it's just been sitting here since then. I mean, I've been playing on it, but it hasn't moved since then. So because I

designed it not to fail within the 30 day warranty. So you're

right, right, it'll catch on fire. And the worst part is you work for insurance by day so you know the ins and outs. Oh, yeah.

Yeah. Yeah, you didn't get that from me. So.

So you're you're kind of evolving in your transformer designs, right?

Yeah, a little bit. So this one's probably a little bit more. I would say useful to the general listening audience I would say but you you're juggling kind of went backwards. I started with output transformers and then now I went to power transformers. And, you know, like anybody who's doing anything that costs money, you try and keep cost down. So I started looking for companies locally, who could source me some parts, mainly the steel or the iron, right? The the silicon steel that was that we use, and found a place in the Connecticut area here and been swapping emails with a guy like, you know, over the past couple of weeks and ended up placing an order for him drove down there about a week and a half ago. And Matt, how

long was this beard? And was it gray?

No, dude, this guy was like your typical like, imagine, imagine the I don't know what you would think of as like, pulled yourself up from your bootstraps built your business from the ground up, kind of dude, that now employs like 3040 people in his own transformer shop. And that's, you'd probably have a pretty good description of this, this guy. Super nice guy, his name, his name is Bruce. And he, he and I were swapping emails back and forth about trying to so first off, he's the only one that answered any of my emails, right. So most people, man, if you're a hobbyist, or if you're a small volume, you don't even you don't even get a call back. Like, there's a handful of vendors out there that will even call you back. Some of them are actually pretty big. Like, I will say Hammond electronics, believe it or not, they eat I've emailed them on multiple occasions about, hey, what kind of powder coat do you use on your chasse ease. And they within the same day, I got a response. So like, there are some out there that will do stuff like that, but for the most part, most people don't have the time of day for you. But this guy, you know, Bruce, he and I, that was kind of the first sign that he is pretty cool guy as he started, he actually replied to my emails. And then he agreed to actually sell me small volumes of supplies. So I went down there, grabbed some bobbins you know, about 4050 pounds of various kinds of laminations. And, you know, a couple other little knickknacks here and there, you know, things called like keeper laminations, which keep your, your stack together and other stuff. But the cool part is, is I'm carrying this big ass box of, of iron out the door, and I was headed to his office and I was going to, I guess give him my credit card number or whatever. And he turns around, he goes, Yeah, you know, it's, it's pretty cool what you're doing, I like I really appreciate seeing somebody, like, follow their passion and really try and start to do something, especially here, you know, with, you know, most people these days are trying to get out of college, go get the corporate job or whatever. So why don't you just like forget about paying for it this time and like walk out the door with $200 worth of, you know, stuff or whatever, which, you know, is no, no drop in the bucket for me. I'm sure it's not for him, either. You know, he's a small business owner trying to, you know, make make, make payroll, and help help feed the families of the people that work for him. But I thought that was pretty cool. Like he and I kind of sat there and talked about, hey, one thing that stuck out to me was he's like, Hey, I had a lot of help when I was getting started. And there's not enough people doing stuff like that for the small guys. So you know, this one's on us, essentially. So yeah, heck yeah, man, I got a bunch of steel and I'm ready to start winding schedules. been kind of crazy here with with kids just got out of school. So I'm juggling all that stuff. And

you got to put them to work then.

Yes, I'm juggling kids. But I like Parker's idea. I think my son could probably wind some transformers. Yeah, there's no AC in the shed, but he'll be fine. Well, he's young.

What was that? What was the transformer shop name?

So it is it's a place in in Durham, Connecticut called technology Trent or let me start over transformer technology incorporated. TTI and the guy's name was Bruce so if you just like I think it's like, transformer technology inc.com. And if you just Google it, you probably end up finding them so they do all kinds of stuff. Like I said, he even quoted me for one unit right for of this particular if I would wanted to get it wound at his shop. So who I don't want to advertise for me like I don't want to say like I go to him he'll he'll do one unit stuff. He might not for your particular situation or whatever. But you know, he's he's a he's a pretty reasonable guy. And I'll say he's got my business as far as anything I need from supplies to winding a Transformers here, you know, in the foreseeable future, for sure.

So what makes a your power transformers different than your output transformer?

So in the sense of, so it's frequent, I would say the main thing it boils down to is frequency response, your mains voltage is 60 here in the US is 60 hertz, and it stays at 60 hertz. So that's that's like, almost becomes a fixed variable in a sense, which, on an output transformer, if you think about it, you're it's the entire audible range, right? Not to mention, not to mention, like, the, the note that you hear, but the harmonics of that note that are coming through that, like, are complicating the sound that you hear but you can't really tell you're hearing right, so it's got to be able to really produce

and I shouldn't interject here real quick. I did test Ross's output transformer all the way to five megahertz. And, and it it had resonance peaks in between zero and five megahertz. But no, no, it totally, it totally shot five megahertz through it, it is it is flat to about 120 kilohertz. And then it starts to wiggle around a little bit in the frequency response, but it stayed pretty. I mean, like I was passing up to five megahertz, like quite visibly on the scope like not having to zoom in

on what about on the other end of the frequency spectrum. Oh, on the low end,

I think it started to roll off at like, sub 30 Hertz. I don't remember exactly where it was. It was low. We're gonna Josh looks smug. I'm so proud. That's pretty good. I gotta gotta get that really good. Sounding five

mega. So for an output transformer, you want that you want a really flat response?

In Latin wide, yeah, flat and wide, right? Because you want it to be as in most cases, some people would argue with this, right? Some people, you know, want their output transformer to at least think they want to call it distort, or they think they want it to start like saturating, which, in reality, I would argue they don't really know what they're asking. But, yeah, in most cases, you want it transparent. And not to color the tone, especially in modern types of amps that were, you know, Steve and I are, I would, I would say mostly messing with most of the time.

Well, and also one of the tests that I did was a the transformer I had in this amp previously, I think I paid $230 for it in 2007. So you know, whatever that is, now that's more Yeah, let's just put it that way. So this this transformer has a total of five wires. So I just unsalted those five wires and sort of Josh's in like directly. But I had done recordings before and after and I did I did a form of recording called reamping where you record your dry guitar signal. And then and then you play your guitar signal back through the amp changing one variable so so it's not me playing differently every time I'm playing the exact same signal

through each time you like pump on the raw signal from your guitar back.

Yeah, the raw signal directly through and and when Roz and I got together and we we listened to both the pre and the after change of the transformer, and I can't tell this quote high end transformer that I bought from some secret sauce place and the one that seemed crazy and the one that he did it is is Shin

Bet you that really fancy one was wound by

some guy. That's true artists and it's I produce artists in Transformers. Yeah. Boutique, boutique. Yes. I mean, in reality, half joking. It does give me kind of flexibility to when you're building one offs, you can do crazy things like okay, what color do you want? You're in bills? powdercoated Oh, metallic purple. Great, we can do that. No problem. So, you know, I'm not saying I'm starting a transformer business or anything. Like I haven't even really seriously considered that I'd say just because of the margins that would be required. But you know, for my personal stuff or Steve stuff or friends that want stuff wound or if I'm repairing an amp and you know, it needs to go

friend him on Facebook. Yeah. Wider transformers

for Empire amp works. Yeah, on Facebook, Instagram, Twitter, we're not super active.

Plug it everyone.

You know, okay, so what really the biggest thing is, is you know, Josh was getting into this, because he and I've been designing some stuff together for some fun. And one of the one of the roadblocks that two of the roadblocks I should say that we're kind of running into is the size of the transformer makes a big difference on the layout of your circuit because these things are huge. So you have to kind of be able to tweak some knobs there. But at the same time, there's a lot of modern techniques that require voltages and currents that are new, you just don't find on an off the shelf trends.

Yeah. So here's the interesting thing about about the guitar amplifier transformer market, I would say is that if you think about it, it's just like any other market, these producers, it's either going to be highly custom, and you have to order it special, or you have to order in minimum quantities of dozens of units, probably, unless you find a guy like Bruce at TVI. Or, you know, you're you're building you're selling to a mass market and your your product is going to fit that mass market. Well, most of these transformers have a specific application, and it's for guitar amps. So they're winding them to either be put into

vinyl or whatever, right classic very legacy,

or people want a specific oh, I want to build an amp that's got that vintage tone to it. So what are these manufacturers going to do, they're not going to wind transformers that don't sell so they're gonna wind this stuff that sells. So if you want to put like DC filament, tap on a transformer and have a DC winding to heat your tubes with DC voltage instead of you know, AC coming off the straight off the transformer which induces home and 60 cycle hum and noise to your circuit, which is a problem with high gain, because it amplifies that home, right all the way through the circuits you have a noisy just hum box? Well, there are a few and transformer manufacturers that do that. But, you know, they're few and far between. And typically these transformers are trying to be they suffered from the same problem trying to be as mass market as possible while still cramming in a couple of these features. So they end up like 15 pounds. And they're huge. And they produce, you know, 500 volts point to point. So it's just a senator tab. So you can like, try and get this massive, do everything transformer. So we, what we were trying to do is get something that was very specific to our application. So that's what really pushed us down this road. And mindset, and

a lot of times in a lot of times you'll you'll have a circuit and you start to design whatever circuit and you realize that, you know, maybe 12 volts isn't gonna cut it, maybe 50 volts is just better for what I'm going to be doing or something in the in the range of 50 volts? Well, that's not that's just not normal, like where you're going to get that you're going to, you're going to have to either step it up with a switch mode supply, or you're going to have to have a nice tap on a transformer. Or are you? What, what have you need 15 volts, but you need five amps off of it. Once again, are you going to put a switch mode in there? Or are you just gonna get a tap on a transformer? And you have the option of buying like six or seven different transformers that all have the tap that you want? Because finding a transformer with a 60 volt tap or a 15?

Not hard.

It's not hard. But you also don't want your project to have 10 transformers and on the top right, yeah, it'll weigh 150 pounds. I

mentioned the cost and like you incur in buying and paying for someone else's profit margin for every unit that you buy

each one right. Yeah, so just make your own.

So just make it and then you get everything you want. And then you get to spend four times as much.

Yeah, so cool. So you're gonna be wanting a power transformer sometime soon.

Yeah, we got here. So

circling back to that, what makes a power transformer different from the output transformer? Right? Yeah, so output transformer has a flat response over the audible range. The power transformer does. Yep. Yeah. Well, sorry, the output transformer. Yeah, what does our transformer because technically, then, in my mind, you could just slap an output transformer as the power transformer.

Well, so it's got completely different windings typically, right? So you're, you're you're trying to match the impedance with the an output transformer whereas on a power transformer, it's all about voltage ratios, right? So you're trying to step usually up from Wall voltage to your main you know high tension or B plus rate which is for most guitar amps somewhere in the 300 to 500 volt rec, usually written that's rectified range. So what your so to answer your question more directly. The main thing that power transformers you want them to do is be able to handle the power. So heat current rate all of the core losses that go into that transformer you don't want it to get too hot and fail because you have some, you know as power as he goes up, no resistance changes, right. So that's going to affect the regulation of the transformer, which is the other thing that I was, I would say is you want a transformer to be able to do is that regulation, right? So if I have a 6.3 volt, heater tap that heats my tubes on my guitar amplifier, I have a 10% wiggle room, you know, no, plus or minus 10%. So, you want it to be able to stay within that range. It's not that particular tap, maybe it's not a great example, cuz it's not super critical. But what if we have like a, you know, a DC circuit that that's designed to do something very specific and like, it has to stay within a tight tolerance? Well, you don't want your you know, you don't want your your tap on your transformer to start to sag or go crazy directions. So you want to be able to, like have that tight? I think the term is regulation typically is what I've heard it called to where it can you

know, the thing is, that's a really confusing. If you've ever spent any time like digging through transformer data sheets, they're not particularly clear most of the time. And that was always a confusing one for me is you get a tap it usually there's there's almost always two numbers, you see what that rep, voltage and current Yep. And that's usually what you get, if you're lucky, you get regulation, which is presented in a percentage. Yep. And then you get, sometimes you get some more stuff, but most of the time, it's just voltage current. And that voltage is supposed to be the voltage that you'll see the RMS voltage at that rated current rate, which means that if you have it unloaded, if you just basically plug it in the wall and look at read the taps, it's going to be higher than

in some cases lot higher, right?

Well, it depends on the regulation, because the regulation percentage is the difference between the loaded and the non loaded. Right?

Right. So going into that, right playing off of that you have to be cognizant of, like I said, heat, right, so the, there are certain things, the core losses from the actual steel that you're using will affect that, because they they actually, you know, for lack of a better term, get excited when you start passing current through it. And that excitement starts to you know, create heat, right? Like similar to like a resistance like a resistor would, right. But then you have you know, so you got a spec your wire, also, because you've got that DC resistance of windings. So it's, you're trying to hit a certain ratio between your primary windings and your secondary windings. And that ratio is what determines, okay, you know, 10 to one means if I pass 10 volts through it, I should get 100. Right on the other end on the secondary? Well, now you're winding a whole lot hope copper wire, which carries its own resistance, which creates more heat, so there's a balance there. But then the other constraint would be, you have a limited amount of space, you can only fit so much wire within that window of that transformer. So it's a balancing act, you have to kind of pick and choose, I would say less of a balancing act than than an output transformer would be and much less subjective. Like you either hit your voltage or you don't like it that's kind of up to the specs of you know, whoever's ordered the transformer or whoever SPECT the transformer, if you want to present regulation on this tap, like, like it might come at a cost of something else. But you know, it's it's, it's typically a balance between all those constraints and less objective than then kind of the output transformer world I would say, which I'm still much like still learning very much a student.

Well, and it's also paired transformers tend to be a bit more robust. Just overall

Yeah, this particular one I suspect is around 260 volt amps. And that I think it's 12 pounds is what it's come out to. Not not not a small piece of mass not it's a big chunk of iron sitting on top of the of the transformer. Now it's probably a little bit over Bill I'll be honest, but that's partially because I was trying to hit that regulation like we've got some we got some windings on there that we went to do some specific things and you know, we we wanted to make sure it was tight enough tolerance for our application.

Cool So when do you think you'll have this all wound up?

Also Fourth of July is coming up here and I'll have a few days off work. I think I might might take a stab at it there. So

I this is what I love about this. This podcast is like most people will be like Fourth of July is coming up. I'm going to the lake and we're going jetski cooks and Barguna Why did Fred

that's what my wife would like me to do. So I have to kind of like sneak off in between

things. So while you're juggling your kid you're also juggling your wife right?

100%

I got I got for you, Josh is cook a brisket, because then there's several hours of

15 hours of inactive time.

Whenever she bugs you just like I'm cooking a brisket.

I actually so it's funny. I actually am going to a party on the fourth. Late in the evening, and I'm supposed to bring a pork shoulder. So yo, excuse built in your genius parks.

Cool stuff. All right. Do you have anything more you want to add to that?

No. Are we No, I would just say I just, I don't know. Just like most other people who are passionate about hobbies or things that they're interested in, I can talk for hours about this. So you know, I'm more than happy to, like I said, answer questions or just kick the you shoot the shit with anybody who wants to, like mess around with guitar amps, or talk transformers, like, hit me up. And I'd be more than happy to talk. I think I'm even on your Slack channel, even though you're in our Slack channel. So, but yeah, that being said, I appreciate the invite and getting the chance to come and talk about something I love. So all right.

Well, we're not over. We're not done yet, Josh. No, I'm still gonna be here. We are here for

those two old guys from the Muppets like firing in, like comments now about things that I don't understand. So

I'm not even one beer through right now. So we've got a little ways to go. Oh, we

get we got plenty of time to go.

Okay, so I've got an update for the Doom should he add on? Yeah, I think I've shown the video of it working to you, Josh. So you kind of know what I'm talking about. Right? Yeah, I do. I do. Well, LCD screen, I've got Doom guy looking angry on it. Yeah. And so over the past week, I got the EEPROM interface to it working, it only works with single byte reads and writes. And I think at that point, I'm just like, it works. I'm just gonna, I'm gonna put it into its documentation, do not try to do multiple reads or writes because it just won't work. Nice, it will fail gracefully as in, it just acknowledges that you did it. And it just nothing happens, of course. And so you can basically what you can request, what, from the first couple of addresses of the prom? simulated the prom, I should say it's just a flash location, right? of memory. You can request like, what kind of shitty add on it is who made it, what year? What DEF CON year, it's four, etc. And then you can also write to a couple addresses. And there are some secrets here. Yeah, so this is going to become a secret stuff that's going to happen. So right now the videos I've shown of it, it shows it cycles through all the phases. That's not what's going to do when you first plug it in. So when you first plug it in, it's just going to show this the neutral icon, the neutral Doom guy face, where he's just like, normal looking, and he's like looking around like that. But then there's gonna be an address in there that is damage and anger. And so the more damage she takes that you can increase the value in the damage register, right? And he gets more like bloody. Or if you increase anger, he gets more angry looking.

Wait, wait, wait, wait, wait. Can I Can I take a guess real quick? Is it is it is the address? I haven't come

up with the addresses yet. Oh, come on. Because

zero h should be zero h 666. It should be the address. right to right.

I don't know. We only have one byte for addresses. Oh, yeah. Well, shit. It doesn't support it doesn't support two byte addresses, or what's called extended EEPROM. Read Write addresses doesn't support that. So that is good idea.

Okay, so H E one one, right. So the the all of the all of the addresses for there's four addresses H E, one and one right. So we go address Hill.

And I still got to make sure that the GitHub goes public. I'm so cleaning it up and adding documentation. Because when I make that public, I want like people to actually implement this kind of stuff into their badges. And I've already had a couple of people come like asked me, okay, I want to put in stuff in our badges so that you plug into doom. It does stuff. And so this is kind of like the beginnings of the functionality of This interface for the Doom ensayo. Oh yeah. And like part of the essay to have shitty add on interface is like just GPIO pins that you can toggle. And so one of them, when you toggle it, it changes what the Doom guy is doing. So like he'll normally this be looking ahead and then when you toggle it, he'll like look around. And then you can cycle through how he moves his face and stuff. So like, that's cool. One person's making like a musical badge. And he wanted to move to the sounds. And so like, yeah, we'll just make it so that it reads I pin and the pins will go transitions from zero to one, it changes the state of him looking.

Oh, gosh, it would be so great if it played the the first level sound. If you plug that in, you know, like the music from the first

way. I think he listens to our podcast. So hopefully it takes that idea.

Oh, that would be awesome.

Yeah, that one's free.

So it's been pretty exciting working. This one's really kind of challenged my programming abilities. It's been a long time since I've done embedded programming. And like building that EEPROM interface. The fact that there's a lot of examples, especially because I'm doing it all Arduino. So it's like, oh, Arduino, blah, blah, blah, site and see, you know, whatever. But the fact that there's no one really has like a simulated EEPROM interface that you can just get. And so I think that's actually one thing that people can get from this project is like that, that simulated IE, like a microcontroller faking being an EEPROM. So you can basically like, shove whatever data you want into, onto the I squared C bus is something that no one's really tried before, or people have tried but not published their code.

Right, right. Yeah, you know, okay, so it's funny enough, that sort of brings up a little bit of a topic that I've been kind of struggling with the past couple of days. And it's the idea that like, for, for stuff like what we do on the podcast, where, in general, like Parker and I were doing projects, before the podcast existed, like we were doing a lot of this stuff, without the need for having it done on a podcast. But now we have this. So there's also kind of like a little bit of an extra drive to do things that are off time. But to have sort of variety, and also have like projects going all the time, it's really difficult to approach a project and have it done in a very professional manner. But also have it done in a short period of time, when it's not your full time job. And I'm not complaining about this. But in other words, like there's this there's this really like there's there's sort of a drive to kind of Arduino everything because you can just crap out code and just work. But I don't want to do that. But I also don't want to spend two years on a single project trying to get it to this like really professional state. Well, like there's like this Goldilocks range right in the middle where you get a good product. And it's not just relying on libraries that someone wrote for Arduino, you're correct, because like the screen interface

is is a library I found, I didn't have to write that protocol. It uses the SAM D 20. One's DMA, memory controller. And that all works. And like I had a writer render, but I'm using like stock functions, net graphics libraries, like the add for graphics library. To do so I didn't write the low level stuff. I wrote a like intermediate. So I can basically shove in a byte array and it craps out Doom guy, right. And I tried to make that that function expandable so that people can basically dump their own own byte arrays in and then it poops out whatever they wanted the, you know, display.

Right, right. Well, I guess what I'm getting at is, is it's kind of it's difficult when you're coming up with constantly coming up with projects where each project could easily eat up a full time job for months, you know, and you and I are pretty good at that. In fact, yes, Roz's to Ross is really good at this also.

Like how many how many irons? Well, yeah,

yeah. How many irons are in the fire?

Yeah, too many, too many. Yeah. Too many to do the way that I want to do them or the way that I see it in my head. And that

is true. That's one thing I've been taken away with this doom SEO is like, okay, sure, I could spend all the time building a library that will work with this screen or I can use something and leverage something else that someone else has built. Right Yeah, and Not and I'm like, okay, the whole point of this project was not to build an St. 77, or St. 7792 LCD interface. If the goal was to build Doom Sal, right? So you had to, I think you have to start looking at your end goal and being like, okay, is that code I'm going to borrow or this piece of hardware I'm going to borrow? Am I going to really learn anything? If I do that process, most of the time, like, I've written code that builds a display interface. I've done an FPGA. So at this point, I'm like, Oh, I can just borrow that code. I'm gonna do it now. Because I won't learn anything different. From right, I might learn a slightly different protocol at this point. So I'm totally fine. Doing that,

you know. Okay, so that totally reminds me of Hackaday article that we talked about a while ago, where it was like, What are you asking yourself to do in order to accomplish a project? Like, you have a project and it needs a piece of aluminum for a jig? In order to complete that project? Well, are you going to build an entire CNC to mill that? That piece of aluminum? And do you need a 3d printer in order to print a 3d printed bracket in order to build the CNC in order to mill the aluminum? That's just a jig for another project? That level? Like, then are correct, true? Yeah. Correct. Those kinds of things actually happen? Yeah, no,

I happen. I find myself planning that up until the point of actually, like, I'll plan the whole project, I'll know exactly what needs to happen. And then it'll get right up to like that phase where I start buying things and then like, wait a second, like, this is gonna cost $5,000. I can't do this.

But no, that's the reason why I went with I bought a 3d printer, instead of building a 3d printer. Yeah, because I'm like, I know, I built XY gantries before and coated them and stuff. And I'm like, I don't need to do that again. To make get a 3d printer, right. So I'm like, I'm gonna buy a 3d printer. Right. And I think that's, that's one thing to look at is you have to weigh all those options. When you start going into your next project is like, well, how far do you want to go? Like, do you want like, let's say, I'm going to put a new engine in my jeep on my jeep, right? Do I want to go to the point where I'm digging an ore mine to get to get the iron ore and then smelting a block? I mean, more power to you if that's what you want to do.

But I mean, that's kind of what Ross is doing. Like he

is, I think the next time he comes on, he's going to be like, folding the steel over an anvil and making that laminations perfect.

Damascus steel. Transformers? Yes,

I can. No, I can see like, he will have hand like hand chiseled a die. And he's pulling copper through it. Yeah. To give him his hand, right. wire diameter.

No, I feel like you guys are trolling. But it's only funny because it's mostly true.

No, I'm saying I admire this. No, no.

Yeah, cuz, you

know, they used to say ADHD was an affliction. But you know, it's things like, like, I don't know. I actually causes me to do some of this stuff, right?

Renaissance man disease that night is

honestly Yeah, it's, it's the opposite of the whole. I don't know, like microwave generation that I've kind of grown up with as a kid, right? Where you don't do any, you don't know how to do anything. You don't know how to change your oil. You don't know how to build something you don't know how to. Like, something that's hard and not your forte? Like, do something that makes you curious. Just because it makes you curious. I don't know. It's kind of like, Why do why am I building transformers because people said nobody understands why what like they say you can't write. That's kind of why I wanted to like, but it doesn't seem that hard. So just go Google a bunch of shit and put a plan together and you're you're now you spent a couple of $1,000 on something.

A couple 1000 to build a $200 transformer.

You know, I think I think it's less about people. Knowing how to make something it's more about people not being willing to learn, or maybe not even not being willing to learn or is like they've just never been taught that you can teach yourself. Like I'm thinking that no one figured

out sieving I think that's the big one right there is people don't realize that they can't self serve themselves.

Or they can well they don't know they can. Yeah, they don't know they can Yeah, like it's it's all at your fingertips. It's it's up to you like nobody's gonna knock on your door and be like, Hey, I heard you're into this. I'm gonna teach you how to do this. Yeah, especially You go figure it out, you

literally have the like the all of the knowledge of the entire universe. On a little computer you carry around in your pocket and you use it to play Candy Crush Candy Crush.

Birds, what's popular game nowadays,

and troll people laugh like crap about political memes on Facebook with your grandparents or whatever. So it's, it's insane. How I mean, to some extent, it's like, it takes time it takes effort to do this stuff, right? So there's gotta be some kind of curiosity like base level. If you don't care about it, you're not just gonna go do it.

So you're never going to learn about how to make a transformer.

And you know what? Transformers six months ago, I would have told you the same thing. I'd have been like, You're crazy. I am never doing that. And here I am. So it I think it all has to do with, right, it's attached. It's attached to the interest that I have in music and guitar amplifiers. And it's just another kind of facet of that. So

we're all turning into old men. Now where else? whippersnappers anything?

Well, no. Josh was talking about his generation, which is our generation. And, you know, people saying you don't know how to mow a lawn or change oil, whatever.

Yeah. Our generation when people tell

me that I like I like, look in my garage. And I'm like, you want to go weld? When you're weld right now?

Right? I guess I guess some of it is that chip on the shoulder attitude, because we're used to hearing other people talk about like, maybe our generation that way, like, oh, the millennials, right? I don't know how many times you hear about people kind of use us as the scapegoat for something. It's like, I don't want to be that kind of stereotype. Right? Which I'm not saying we are. But that's like, alright, well, I wind freakin transformers and do crazy things just like you're building a do shitty add on. Like, who goes through that level of detail for something that's, like, just kind of fun, right? I don't know. It's, it's, it's hard. And it's not the everyday thing that you see people running into and doing just for fun. So well, on

top of that, I think, I think Parker, you kind of nailed on the head right there. Like with what you were saying was like, Do you want to come over and weld? Like,

I don't think Parker is the kind of guy that's going to, you know, just open up his garage and have a big sign on or it's like free weld shop or? Well, but, but like, if somebody came up to Parker, I know in general, like if someone was like, hey, I want to learn how to weld Parker would be like, Okay, well, I'll

show you some things back up here in front of my house at 10 o'clock in the morning on Saturday, and weld all day.

But the biggest difference is, you have to ask six, you have to be the one who's like, Oh, you have a welder? Can I learn? Like that's all you have? To do?

Exactly. You have to be you have to take that first step. Right.

And bring a six pack of beer in it for over 20

ones. And that's the same, like going back to the story I told about that.

Have your dad bring me a six pack of beer?

Six Pack, I'll get you through breakfast. But what are you gonna drink for the rest of the day?

Well, you also wait but the thing is, you also have to be willing to except that Parker might sit in a lawn chair and drink that beer and make fun of you while you will. That might happen. But that's also like go that's part of the training. You know, Pat Awami these

guys that from welding. You want to rack up these guys from?

Oh, yeah, that's our old man rant right there. Get off my lawn. Okay, I guess on to the next thing, which is me, right? Yes. Unless Do you have anything more to add on to the Doom shitty? I

think we're well off topic by now. Yeah. Okay.

So a quick update. All the parts for the macro amp version 2.0 are on order. Once again, thank you, Mauser for all the help you've been given us on that really appreciate it. We will have some cool updates on that. And I've got some articles coming out. That's going to talk about the entire design of it. So if you want to follow along, and watch that even the construction I'm gonna have a whole thing on like, here's all the design choices I've made in order to make you know this thing. So that'll be that will be fun. I am going to have to do I bought actually a new enclosure that is a wooden enclosure. And just because of the thickness of the enclosure, nothing can mount to it because it's so damn thick. So I'm going to have to do some goofy crap where T H ICC? Yeah, yeah, exactly. Like Like I want RCA jacks connected to the enclosure itself and I want them connected to the wooden part not the not the top part just from the way it looks. But I'm going to have to use a forstner bit or a paddle bit on a And hog out wooden material to make the wall thin enough that an RCA jack could actually screw to it. Yeah, so it's because I'm like, I love most of the design of this. That's the one thing where I'm like, gotta hate that I have to do that because it's gonna look not so great. And it's also like sort of really like goofy custom but whatever. Sometimes you have to make these compromises. What about being engineers

about

like a recessed panel, like you could route out like a panel and mount like a piece of aluminum or something in there. Stop being

smart Roz? Like, just just be mad with me about my decision would the recessed panel thing is is a good idea. But it's more work like the pedal.

Right? There's like, yeah, a forstner bit of power, but you just like, Yeah, to draw the dot and you're on a drill press. Whereas now you're talking about jigging it into a scene. See, you've got to build a CNC. First you gotta you gotta build a fixture to hold the

case. Next week on the macro fab podcast. Steve starts to build a CNC

Oh, next year.

Yeah, Steven has a CNC actually.

It is not built currently.

It used to be built. It's like a, it's a four by eight man out of wood in a chain drive, which is ridiculous.

That always made Barker mad that it was chain drive. Yeah, when I cook guitars, it worked. It did work. And eventually I will rebuild that. But you know what's funny? I don't have a project that needs it right now. So I don't. Yeah, no, no, I know. Rosin, I have some projects that are in a cauldron that are boiling right now. And that one, as soon as they spill over, I'm going to say, Well, shit, I need to see. Let's see. Eventually, that'll happen. I'm trying to avoid it. I mean, you guys can see my basement right now. There are things to do in here. So yeah, you know, actually funny enough with back to the paddle bits in the fortune of bits. You know, what the biggest thing that I just don't look forward to what that is. So Parker's absolutely 100%, right, you just mark the center, and you just pull the lever on the on the on the drill press and you get a hole with a flat bottom right, right, biggest thing that I don't I really just don't want to do right now is set my depth. I don't want to find what the depth is, and have to set it for each hole individually and do all this crap. I just don't want to

go buy a $99 drill press from Home Depot. And they have DEF stops on them.

Yeah, exactly. That's the part I don't want to do. Now the funny thing is like I run a CNC all day at work, like I'm in charge. Well, in fact, now I set up a second one. So I'm in charge of our entire metal milling department. And it's funny because I love doing that. But I do it for eight hours every day. Now. I kind of want to come home and not do that anymore. You know, I get to it. So don't get me wrong. It's it's a ton of fun and whatnot. But just like, I'm not looking forward to that. I don't know why? Well, actually, you know, here's the thing, I designed the second board for the the macro and version two, I designed the board. It's a single board. There are very few solder points. It just drops into the chassis and everything has screw terminals. So all the Transformers connect directly to the board via terminal blocks. So there's like the the actual construction of the amp is really quick, which I'm looking forward to.

So the chassis, is it one of those like Hammond chasis with like the walnut trim on it. And those are always those always look nice.

Yeah, yeah. In fact, I will hold it up here so you guys can see it. I've got this is. This is the version one. Like man, yeah, yeah, they're really nice.

Wait, what's that iron on the inside? Is that a choke?

Yeah, yeah. That's a monster. So I don't know. I think it's a five Henry choke that's in there. So I ended up scrapping that and making a solid state choke Gyrator design, which we talked about in the handful, yeah, episodes ago. So that's going to go well not it's already on the PCB. So a lot easier to design and implement. And I even went with PCB amount of tube sockets. So everything just kind of slaps right into the chassis and goes, that's sort of why I was not wanting to do the whole chassis modification to get it done. But regardless, whatever. That's all that so all of those parts are on order. Should have that in about a week or two when macro fab decides to build my stuff and send it to me working on it. Yeah. So last Episode I talked about since Mac ramp is basically in a close to done phase, I wanted to move to a different project. And I wanted to do something with a processor that I actually have to, you know, design and write firmware for and do some coding. So I've been mulling over ideas. And this isn't set in stone, but this is sort of one that's come to mind. And it actually solves an issue that I've been wanting to solve for a while. This is

this is this project falls. I just thought it falls into the the current theme of this podcast as well. Of how far do you want to go back to actually building stuff? Oh, you can buy one of these things?

No, you can't. Are you kidding? Okay. You apologize, you can buy one of these things, but it has, it has one critical function that I've never been able to find in anything I can buy. And I want to discuss that a little bit. And funnily enough, it's a really easy function. So I want to buy a fridge controller for brewing, brewing beer. So in fact, not only can you buy one, I actually already have one.

You have a st was sec 102. Johnson's

control. Okay, you

got Johnson? Because I have the inkbird. Yeah, I got one of those. Yeah, he's got inkbird. I've got the STC 100. Which is that little module that you just like, yeah, yeah, like, I got hundreds of those, well, maybe not hundreds, but lots, like two. I mean, they're like 20 bucks, I have like, you know, I have like five of them?

Well, okay, and here's the thing, I've had this, gosh, wow, I've had this same controller since 2010, I think is when I bought it. So it's been going strong. And it's been running. Not entirely that that whole time, but for a good chunk of the last nine years that fridge controller has been been running. And what this thing is, is it's basically just a relay in a box with a brain and it has a temperature probe on it, you put a setpoint on that temperature probe, when your fridge passes that threshold, it turns the fridge on the fridge can cool down, and then it turns off, that's all it does. And that's fantastic. And it works really great. There's a there's a downfall in my mind with that design or two, I should say, first of all, it has only the capability to cool or to heat not either. So I want to design one that can switch on a fridge or it can switch on a heat lamp. When I lived in Houston, I only ever needed cooling. But up here, I need both it actually it does matter. So I want to be able to have a threshold and, you know, be smarter than the thermostat in my house and be able to switch from heat to cool, which that's something that we complained about the other day on the podcast. The other thing is, and this is where the I've never found a controller that has the capability, I ferment at any one given time, I typically ferment two different batches of beer, I want to be able to have probes in both batches. And I want to be able to take the average temperature between both of them. And I want to be able to set the fridge temperature to the average temperature of either brew. Because sometimes one brew can be going way more vigorously than the other one. And they can actually be producing more heat than the other one. And so you can have the temperature of the one that's not as aggressive be off, if you're trying to measure the temperature of the aggressive one or the other way. Basically, I want to be able to monitor both of them. But on top of that, I also want to be able to monitor the ambient temperature. So I want thermo wells that go down into my fermenters, where I can have temperature probes that are measuring the actual liquid temperature. But I also want to measure the ambient of my fridge and I want to be able to get an idea of all three of these temperatures and be able to make a more complex algorithm, as opposed to just like, Oh, I've reached a threshold turn on or off, you know, sounds like

a glycol system, bro.

So yeah, I was actually thinking about the glycol system, because I had an idea to build my upright freezer from my from a tater. Yeah. And that's the biggest problem is because a lot of times I'd have two or three beers brewing or fermenting in there. And you're like, Well, I kind of had to pick one temperature to keep the whole fridge app. Right. Right. And usually you pick like the freshest the newest beer in there is the one that you set the temperature to. Right, and it goes after seven days. It really doesn't matter what that beer is set to anymore.

No because it's done. It's done.

So I was thinking like, hey, it'd be really good if you could put up thermal well on all of them, right? And then all them have their own, like a like a like a blanket around them. heated blanket, because you can get like firmer tighter blankets that go around that heats them have that and then you keep the fridge to lowest temp the of your beer. So like I say you're fermenting it at 65 degrees Fahrenheit for one of them. And then all the other ones have a higher setpoint. And they run to blankets more.

It's kind of like, energy efficient, but I like it. It is

very energy inefficient, but it would work and it would be a lot cheaper than a glycol system.

Well, the inverse of A glycol system instead of like, yeah, glycol, basically just, it's the opposite. You put the tanks at room temperature and you're just sending cold glycol through them. Yeah. In this case, you're putting them in a cold chamber and sending warm goodness around them warm electrons

around them. Yeah.

Well, that kind of what that thing you could you could technically lager and Ale at the same time? Yeah. Which, which would be pretty cool. No, no. Another another feature that I'm kind of thinking about adding to this controller is, so fermentation is an exothermic reaction. So when the yeast is going nuts in the beer, they're actually warming things up. What could be cool is instead of taking the average, you could just say, make sure that the hottest one doesn't matter which one it is, but the hottest one never goes above this temperature. So you can have a global threshold in the fridge for whichever one is the hottest. I like that idea, too. So that's how I used to run mine, but I only had one probe. Right? And I just put that probe in the whatever was the newest beer. Yeah, so you're sort of you're sort of tricking it. Like, well, you're sort of just assuming whichever one I want to design this. Yeah, yes.

Yeah. If you, we should combine these two ideas, because then you can design it, and I can just buy it from you.

I like that. Yeah. That's great. Well, okay, so I want to use some DS 18, B, 20, temperature probes. Those are super cheap. They're actually like they're half a degree Celsius accurate. And they have 2% tolerance, I think. So in general, they're super easy to use. What's nice is they're also one wire compatible. So, in fact, I think someone hit me up on Twitter the other day, I can't remember who it was, I apologize. But they had 48 different temperature probes on a single pin on a processor. And they were just getting a huge aggregate temperature data off of it. So in general, I want to I want to get a nice box. In fact, I'm thinking about getting one off to Kouachi or taka chi, I don't remember how to pronounce that. I think Tokaji, they make excellent enclosures. And they will do like printing and milling for you. If you want that even at low, low volumes. In fact, trade German, we had him on once and he was showing off his altimeter I think it was project and he had to catch you make that those enclosures for him.

It was my favorite thing with that company, cuz I actually liked working with that company, because they work with you with like singles or up to production runs. Right? It's to Kouachi sounds like a Japanese company. And they're a Chinese company. Right? Right. Yeah, yeah, I always thought that was that was interesting.

So one of the other things I want to do is I want to add Wi Fi connectivity to it. And that's, it doesn't need it. But it's more, I've never, I've never I've never designed Wi Fi into a product before. So this is a long lines of what you were talking about earlier, where like, I want to do something because I will learn because I'm doing it. And and the AI spindles that I created, they had Wi Fi but I wouldn't call those mine because I just followed instructions, you know. So I want to implement Wi Fi. And pork and I were talking about it the other day, I sort of don't want to use the ESP 8266. And that's not because like I want to be ironic, and do something different than everyone else. Really i The 8266 is just, we've we've discussed issues with it. There's some headache that I want to avoid. And I don't want to just, you know, write some more Arduino code. I've done that before. I want to learn something new. So I've started looking at different Wi Fi modules. And actually, I found that Silicon Labs, which is something that we've talked about a lot on this podcast, Silicon Labs has a lot of solutions for Wi Fi connectivity. They have some good modules and a lot of really, they have some powerful stuff things that are like the Wii most D one mini and a lot of other things of that sort and there's a general ESP stuff. But they also have some really simple modules that are just basically power and serial where you just give it data in and out into the atmosphere and data comes back. Right and to the mothership. Yeah, it automatically it has the mothers SIP stack built inside, right? So red ale. Exactly, exactly. Well, and it's funny because if I go with the Silicon Labs Wi Fi stuff, Silicon Labs also has a USB to UART bridge, I see. They also have all the AFMA processor stuff, I was thinking, why not just Silicon Labs it all the way like do every chip that I can Silicon Labs just for fun, you know, because technically, I've never done anything Silicon Labs other than play their Space Invaders on their dev board that that we got, which well, and I apologize, I modified their code to make Space Invaders run at like 5x speed, and it was impossible to play. That's that's the extent of my Silicon Labs work, which was cool. So in general, that's sort of what I'm thinking about for my next project, I think that will be kind of fun. And I also kind of want to do the Wi Fi as an external antenna, as opposed to the like PCB trace antenna, mainly because I was looking at the Tokachi enclosures, and they have some really cool looking industrial enclosures that you mount to the wall. And I'm thinking I might make this look like something that you would have purchased from an industrial supply, you know, and I don't know that just that sounds really fun. And their enclosures are really well well done their data sheets too, because they, for most of their data sheets, they provide a PCB layout example of like, make this and it will screw into our enclosure. So you can you can do some really nice stuff with that. So cool. That's That's enough of me rambling about my crap.

So that's what we'll go on. Because I'm actually looking forward to it because of you. I want to use that for my fermentation.

Yeah, yeah. Well, and you might cross your fingers or sipping my Mac or hemp works. You might also end up with a Mac ramp at some point in time. Oh, that'd

be cool. Yeah, I'm actually been setting up my Hi Fi stuff here. So the fun. Yeah, we need to build a power amp then

Demetria is a power amp.

Oh, it is a power I thought was a preamp? No, it's the whole thing.

It's the whole kit and caboodle. Ah, you got you just you just plug anything into it and then plug speakers directly into the back.

Just gonna get some new speakers then.

That's right, go down a little thrift store. They always have them.

Build some of those plywood speakers.

Oh, yeah, the exciters on speakers. Yeah. Yeah.

That's another that's another show.

So I guess we'll go on to the RF Oh. So the first RFO is confirmed. NASA has been hacked, done. And so this is a audit that was performed by the US Office of Inspector General was published by NASA this week, which revealed that an unauthorized Raspberry Pi computer was connected to the Jet Propulsion Laboratory, which is JPL servers, and was a targeted vector by hackers use it to move further into the NASA network. So they were able the hackers were able to access all the way through to the Deep Space Network array. They hacked space guys

with a PI with a Raspberry Pi. If you don't know, don't wait. If we went back 30 years, and just said that exact same phrase people be like, wait, what? We have space with a Raspberry Pi. That sounds like something that would be back to the future. And so

the only only thing I really want to talk about on this one is do we think that was the Raspberry Pi just something like some server admin was using? As as you know, it wasn't intentional. He plugged it in to use it as like a node or as a serial terminal or something like that. Or do you think someone actually planted this device to allow the hackers are getting to the servers

now somebody at NASA was goofing off on their lunch break and forgot to unplug their pie. I guarantee that's what happened.

Well, and you know what, you just have to see did they change the password? Or is it still raspberry? And

you're correct, Steven? Yeah, I bet you it was an accident. And then someone, some guy was just pinging, you know, trying to find Raspberry Pi's on the internet and then hit one that was at J had a JPL, your IP.

Or we're sending we're literally sending $35 computers into space to power things which I would not be surprised. Yeah. And they're actually using it for like some kind of development. I mean, JPL is like the cool part of NASA too. They do all like the

robot. Yeah. I say Russian collusion.

Big news.

So I would say, looking at, like, looking at the document and stuff, I think it's probably someone left it plugged in and it wasn't supposed to be plugged in.

Like I said, check that check. If you check the password, that would probably tell you the answer. Yeah.

And the password wasn't Russia collusion.

Russia. No, it's just Well, I guess this week, there was also some other Raspberry Pi news right burger.

Yeah, so the Raspberry Pi four got released earlier this week. It's faster, better, stronger. Uses type C USB for power. There's not a lot of information about that. Like it actually implements the, the USB power delivery service. It just says it has USB type C. So that can be anything has USB 3.0, which is fast compared to the old 2.0 that used to have, it has a lot more RAM. You can get four gigs Now, up to four gigs. Yeah, up to four gigs has gigabit Ethernet, which is cool. And I think the coolest feature about this though, is it has two HDMI ports. And it can drive 4k monitors,

which is pretty cool. Yeah. And they're the mini HDMI. That sucks. Yeah, yeah. Now

owns those. Yeah, that cable. But

it's funny, the it's been slowly progressing. When when when the I think the Raspberry Pi one and two, you could both run off of a one amp, five volt wall charger. And then the Raspberry Pi three was like, well, you can but you're gonna lose performance. And now which is five volts, three amps, it runs like, it requires a 15 watt power supply now.

Yeah. My biggest problem with Raspberry Pi four is the whole point of the first couple was like getting computers in the hands of everyone. It's kind of like why the first versions of these they had a composite video output. So you can plug it into any CRT TV. Right? That for monitor. This is kind of like I mean, sure it's the same price. So it's like it's still like the cheapest version of this. The Raspberry Pi four is still 35 bucks, which is amazing, given like inflation and all that stuff. But who tariffs is has 4k monitors to plug this thing into. I mean, I guess you can use the standard monitors, but you still need these like special HDMI mini or micro cables to normal HDMI, which not everyone has lying around. So you still have to buy those cables.

Well, so I think a lot of laptops actually don't they have. I'm trying to think like they have the mini HDMI. He might have mini HDMI, so yeah, I don't know.

Well, right. But I mean, the if you're buying this, you already got the computer parts, right? I don't know. It's cool. Like it's,

I like how they're improving it. It's just one of those like, that HDMI decision where that cable, it's like, oh, I had to go find that

cable that strikes me is like a space consideration, though they were probably was more outputs on to the board. And

because I kept the same form factor.

Yeah, it's always been the same form factor with the same mounting holes.

No, they they change was the IO pins. Oh,

yeah. You know what changed on this one. And I'm not entirely sure why there is a major change on this. The Ethernet connector is on the other side of the board. Now not fully the other side of the board. But if you're looking at the side that has the Ethernet and the USBs, it's flopped. So it actually won't fit in your cases for the Raspberry Pi 312 Or three, it will require its own case.

Oh, how long will it take for someone to bake a new Raspberry Pi four case on Thingiverse? I can just

it's already it's already done. Yeah, guarantee. Cool thing is you can actually you could do this with the three also. But this supports Power over Ethernet for the full three amp five volt supply, which that's kind of cool. That is cool. Yeah. Although a Power over Ethernet hub that supplies that much power is not particularly cheap. So you would Yeah, you'd need a reason for it.

We're going back to complain about the HDMI ports yet to like Power over Ethernet.

Well, I mean, if you needed if you if you had a need for multiple of these throughout your factory or whatnot, you then that's a really easy you don't have a bunch of Yeah. Well, you don't have that power all over the place. Yeah. Yeah. Cool new Raspberry Pi, just in time for me to not have learned the last one.

So the next topic is the I thought this was really cool. The research platform flip. And flip stands for floating instrument platform. It's a 355 foot long research platform. It's a boat, they say it's not a boat. They say it's a research platform. It's a boat, though. It's 355 foot long. So it's longer than a football field. And for people who not in America, it's longer than a soccer field. Pitch. Pitch. Pitch is a pitch. Okay. Yeah. And I said soccer. Right, great. And so what it does is it starts it's really long. And so when they get into position in the ocean, they fill it up with water as a ballast, and it and like 300 feet of it goes underwater. And so it's a big, they call it a spire buoy. And so it doesn't move at all, when like waves hit it. It's very stable. And so they can do a lot of like, current research and wave motion and stuff like that. But the it's interesting how they designed the interior to handle the fact that when you're moving it, it's looks it's like 90 degrees, and then when you set it up, it goes 90. So like the bath, zero, no, no, everything has everything, or it pivots. That's cool. So like the bathroom has like to have everything one is just 90 degrees out of phase.

So 1962 It's a little round today to

Yeah. Wow. So it's funky looking when you see like, when it looks like a boat is sinking? Because it's like it tips up and looks like it's going bow first into the water. Right?

So I guess they fill it up with water. And then when they need to drive somewhere else. They have to pump all that water out of it. Right? Yeah, probably. That probably takes forever.

Probably. This is one guy with a hand pump. crank crank.

Yeah, well, well, all the other guys are sitting on lawn chairs, drinking beers watching him. Yeah.

Yeah, just don't be on the the bottom part when they hit that flip button, I guess.

Don't get on the Don't be. Don't be on the wrong side. Yeah, be on the floor, toilet

or don't be on the toilet when you're starting to flip. Next topic is the automated dice tester uses machine vision to ensure a fair game. This goes back to kind of what Josh was talking about where like, you find something that you enjoy. And you can just like, focus on this guy wrote a 62 page report on how dice, how the manufacturer dice influences the rolls for the game X Wing, the board game X Wing.

Awesome. That is pretty awesome.

And so he built this like thing that would roll the dice. And it would look at the dice with a camera to figure out what was rolled. And so he can just automate and get a ton tons of you know, data definitely builds

a predictive model on top of that. He built

a machine vision neural network on top of it. Wow, that's so like, it's even using AI basically, to see if that dice rolls are fair or not. And basically, he found out no, they're not. Some dice have, like upwards of like five to 15% bias on some of the rolls. Hmm. And he's like, well over an entire game. Yeah, that would totally make sense. Like you that can influence the outcome of the game and he had good dice. And so I think this is kind of gets to where some people have like Lucky dice, right? Like in d&d, like you have a couple sets, but there's one set that you use a lot, because you got good rolls. It's probably because of this because the manufacturing process of dice makes them inherently not fair.

I wonder if there's any other factors that influence that like going back to your lucky dice. kind of example, like is a dice lucky for Josh because he rolls it a specific way and Steve has a different technique and that's true.

Dice are always unlucky for me. That's one of one you know that way I always loved access and allies because most games positive numbers on dice are mean good, but access and allies was designed that lower number Is mean that things are successful. In other words, the number six on a die six is the worst number, you could roll in access and allies. And it must have just been the guy who created it was just like, super unlucky all the time. And he's like, I'm making my game the opposite.

But it's it. It reminds me a lot of when we play online d&d With rule 20, dotnet, because it has simulated dice rolls, which are probably fair mean, that is a one out of 20 chance when you roll a d 20 of each number. Whereas in real life, you would roll you're lucky D 20. That has some slightly bias to the 20,

which is actually loaded, apparently,

well, they are loaded, go obviously actually read that report. It's 60 pages long, but it's actually very interesting. Because he goes through his method, he goes through his methodology of how he goes into the design of the tests, and what's important, because like, he has the actually the test is reading the bottom side of the dice. Right? And not the top side, but it really doesn't matter because you can just, you know, extrapolate what the top of the dices. Right? So he goes through this whole process of like, the whole design and, and the whole how the game, how the game works, and how if, let's say you had 5% more of this, how much it would actually influence the game. And it's like, okay, that actually is quite a bit more.

Well, okay, so So I got I got a monkey wrench to throw in here. So he has at the at the top of this, I guess page three, he has a certain, here's a chart that shows the number of dice and how many rolls they had. And in some of these cases, 31,000 rolls to get a number that's great, you know, lots of data points mean, you can be fairly confident in your data. However, these dice are not made of Invincible material. Is he cooking in errors by wearing his dice? You know, 30 something 1000 rolls, I guarantee you those dice are not going to be the same as when they were fresh out of the paper. How

many times have you rolled a D 20?

Me? Oh, gosh,

I don't know a lot. Well, like one like,

Oh, I like a single one.

You will have you rolled a single D 20 20,000 times

No, no, for sure. Not 20,000 times, but a few 1000 problem.

So yeah, I wonder if we dig more into this this paper, like maybe has an over time bias?

Yeah. So here's, here's my question. Now, here's my issue with his test. And I haven't read it. And I'm literally just scrolling through it. So yeah, he's got a lot of rolls here. But I don't see that many different sets of dice. He does

have different kinds of dice. And there were two other dice have only seen

a handful though. But my point is, is like, yeah, he's got a lot of different samples from a couple. Maybe it doesn't tops, different actual dice here. So I don't think you can confidently say, dice are biased. Maybe the dice he tested her bias.

And the thing is, though, is he also tests the there's apparently an app for this game. He tests that as well. Like the dice rolls in that. So there's he's pretty thorough, at least for this one game.

Yeah, I'm gonna read this because this is interesting. This is busy.

This guy must have been screwed out of a handful of games or something. Yeah, no. Coffee tables.

Yeah.

You know, I've seen some dice before you Okay, so you know, how like normal dice have divots for the for the numbers on them. I've seen some that are painted. But I've also seen some where the divots are all equally weighted, so they're removing material from each face, such that equal amount of weight. I also saw one the other day that was was interesting. It was a gimmick dice at some game store I was at but it was a large dice six, that was transparent and it was hollow. And it had a second die six that was inside of it. So if you ever needed to roll two dice six at the same time, you just rolled this one, you read the outside and then you read the one on the inside. And I bet you that one is far less predictable, you know, just because it has this odd maths on the inside that rolls around. Yeah, yeah.

But he tests a dice that is marketed as fair. And he found that some of them are not fair. Because of how they're manufactured as well.

Interesting. Well, and I actually watched a video of a guy talking about a lot of a lot of dice are when they're made, you know, they they they get put into a tumbler Yeah, and that's actually covers that Yeah. And that's actually affects the curvature of the cusps on each face. And that can have a huge impact on.

Yeah, he actually measured some dice and like the be a millimeter and a half of a difference on a side of her face,

which is actually pretty big.

So I was kind of thinking down that next is like, Okay, now he's, he's come up with this result. But can he now narrow it down to factors that predict the? Because yeah, I'm thinking through application, right? No manufacturer of dice is going to sit there and roll 10,000 times for one died. Yes, it's fair, that sort of

process is the right

dice is $20,000.

But if a manufacturer can put it on a scale and measure it in a fraction of a second and determine if it's fair, or look at it with a computer, and you know, determine whether it's measured with a caliper, whatever, right, if he can come down to like the does he go through that in the analysis at all,

I think I take away from this is I should build one of these and then dissect it and then pick,

find the ones that are the most bias towards like the good ones.

That's what I'm taking away from this.

So I like it, here's, here's what I would do get this, if I wanted to make a perfectly fair dice, what I would do is I would first pick a material that has an incredibly uniform density, something that's like, like a, like a composite, or a ceramic, or something of that sort, where like, you cut it any angle, and you're gonna get the exact same face, and you're gonna get a very uniform material. And then I would mill and lap every face of it, it would be sharp as hell, but like, the facets would all be as sharp as possible. And then I would laser every numeral on the edge of whatever face it was. But I would make sure that the lasering had the same surface area that it was laser in each one. And that would be my

career, make a billet aluminum, D 20, on your CNC at work?

I bet you we could I the the biggest problem with with things like D 20. Is how do you hold it? That's true. That's so hard. If you had a five axis CNC, it would be a lot easier.

Tune in next week as Steve builds a fiber

and a ceramic forge or whatever. Yeah, I think because I actually help. So my, my sister and my brother in law, were actually selling dice on Etsy for some period of time. And they actually they made some good money doing that. But they were making them out of wood. And they were doing dice six, and they hit me up once and they were like, You got to see and see, can we cut a die 20. And I said, let's give it a shot. So we actually did it. I built a jig that bolted to my CNC and we held a chunk of wood and we milled half of the 20th. And we flipped it over and milled the other half. And it didn't register super well. So there was we had to sand it afterwards, which wasn't a big ordeal, or anything like that. But the problem with a D 20. Is it doesn't have any two faces that are parallel that you can grip. So no matter what with the deeds, what are your fixture, well, you'll get one good face, and then you get one shitty face or like three shitty faces that you have to make a unique fixture that registers on those extra faces on a on a vise, which, okay, not super hard, but also a pain in the ass.

So I wonder The solution is to build a better D 20. With like a different one that you can machine correctly.

A dt 20 that has two parallel faces. I don't think that exists. Like I mathematically I don't think that exist. The thing that's special about the D 20 is that every face has the same geometry. Right? Correct. And it all fits together is that the D 20 is a soccer ball right? No, no it is not. I think girls soccer ball has different shapes. Yeah. T 20 is the one that's just incredibly unique. Because of it.

It's yeah, a bunch of triangles right? So

they're, yeah, they are triangles. Yes.

I can go like icosahedron Casa Heejun icosahedron

or you can do what you can buy online and you can buy a di 20 That only that doesn't have a one two or three on it has 220s to 19 to eight teens. Yes right which is why Steven always made me roll in the in the app. Yeah. I bought those assassin my curiosity so I never I never I never cheated in dungeon dragons. I promise

just Star Wars. Because we play Star Wars.

I never cheated in Star Wars. So sure. I always rolled nap doesn't matter. I wonder if the app is biased. I guess we have to check that out.

I was biased for your character because your character either road ones or 20s. Like that's true. There wasn't not a lot in between.

There wasn't. All right. Oh, man. It's flush. Like, Oh, God. Use computer use roll one. Well, I guess my guy's not going to be able to know how to use a computer for the surround. My guy

who's used computers for everything, including like going to the bathroom since he was a child just forgot where the keyboard was. Yep.

Just no idea that keyboard catches on. Yeah. Okay, so next topic. Last ones, Scott long podcast. Yeah. G lipid bulbs reset procedure. So this is taking the internet by storm this week. And involves a lot of on and off. So this is smart light bulbs, not your typical analog or regular LED light bulbs. And the whole thing was people were like watching this. And it's like, it's ridiculous when you actually first watch it because like, turn it on, turn it off. Turn it on, turn it off. And you do that for like 30 seconds, right? But like

it's turn it on, let it stay on for eight seconds, then turn it off. Then turn it on, let it sit for four seconds or whatever you like. It's the

same, same eight seconds. It's like eight seconds on two seconds off, repeating for like five times, of course. And people were making fun of it because of how a bad like procedure that is. The thing is, though, I think if you had the on your phone, you had that video open, it wouldn't be that big of a deal. Because you just listen to the guy say on off on off. Wouldn't be that big of a deal.

Yeah, they clearly I think people just blew up mash the light switch either.

Yeah, it's like how do ya how do you? How would you implement a reset procedure that wouldn't be used in normal life? The only uses of binary input?

How about a damn button on the side?

Yeah, so whatever it was, like 30 feet up in your ceiling.

That could be a problem.

I would say you have to you have to flip it on and off techno rave to darude sandstorm, and for the first 10 seconds of the song, and I

was hoping you'd go Homestar Runner on me? Oh, yeah, yeah, you're right. Cheat is grounded.

Wow, that's, that's high school for you right there. What are you okay, so here's the thing. How do you know that you reset? Like, does it blink in a pattern? Does it tell you?

I don't know. I think basically, they're resetting the Wi Fi. So it because how these things work is when you first turn them on, they broadcast over their own SSID. And so you connect your your phone to that SSID, right, set your Wi Fi, name and password, and then it resets connects to your home Wi Fi. That's how they work. So you're basically you're resetting it to a state to where it's broadcasting back on its, you know, default SSID. And so it's like, how do you do that? With some a light switch.

Okay, so I think one way that would be reasonable to do that is have it instead of like this eight on eight off, you know, eight seconds thing, have it where you flip it on and off quickly, like 12 times, like on off on off on up, you know, you just hammer it a couple times.

I'll just hammer it until it resets.

I'll send my six year old to your house, and you'll see why that's a bad idea.

Exactly. Because I remember when I was a kid, I would foot light switches like crazy. Yeah,

literally light switch, instead of

12 times make it 20 times I don't know, whatever, whatever. That's

worse than just flipping it for eight seconds on and off for five times. Because then you're sitting there all day, like, you know, I could say something that's probably not very good on this podcast, but yeah. I don't know, what's a good way of repetitive up and down motion for 30 seconds.

Well, okay. How about you just have like a code that you send out from from an app on your phone or something? Like, like a I don't know, like a

train? Yeah, you're right. But I'm gonna assume basically,

it listens for a particular signal and when it hears it, it goes into default mode.

Yeah, you could do that. But the thing is what happens if

If you reset your

modem and you didn't know what the password was because you're using the stock, WPS, you know, push both.

Yeah. Because then people could like come reset your, your lights from the street or whatever if they're just parked close to your house. Yeah.

Well, I don't, there's not really good solution to it.

And there's reason GE is going bankrupt. So,

yeah, they're too big to fail.

Yeah, I don't really know. I think the best solution is not to have smart lightbulbs.

Yeah, I was just thinking to that I literally don't know what a smart light bulb does. Are we talking about the ones that change colors? Like

they do? You okay?

Is that what they do? You know

what? I'm gonna go on a tangent here. I got, I gotta teach it that. Just hang on with me for a second. Do you know, okay, so it's a running joke. It's been there for gosh, however many decades, like, You're a really smart person, if you can set the clock on your VCR, right? Like, how many sitcoms Have you seen where it's like, I set the clock on my VCR and you are like some kind of like God or whatnot like that some kind of difficult thing, right? Sure. I'm not trying to like brag or anything like that. But like, it's not hard to set the clock on a VCR, even if you don't know how to do it, like, you can figure it out. And a lot of times, maybe it takes like a certain mind, or maybe like there's like a way of looking at something. But if you approach a device that has a single button, one button on it, you have some options, right, you can press the button, you can hold the button, you can press the button, like quickly, you can hold it and let go. Like there's only a couple of things you can do. And so if you look at like a VCR or whatnot, there's, there's only a handful of buttons on them. Most of them have nothing to do with clock, so you ignore those because whoever designed it would have to be a psycho to make you press like fast forward in order to set the clock or whatnot. So let's let's let's backtrack and go to the light bulb idea. You have a button, you have the light switch, right? That's, that's all you can do. You can turn it on or off. So your options, position fast or slow or time, right. Those are your options.

And it has to be something that a kid isn't just going to randomly fall into.

Right. Okay, so I failed that one. So maybe, you know, maybe GE is late, even though everyone's making fun of them. Maybe that's like the best way to do it.

I would say the problem with it is they have two different versions of it depending on what light bulb you have. firmware. Yeah. Okay. That's the big problem, though is, you don't know. The only way you can know what firmware you have is by the outer packaging, that light bulb came in? Who keeps packaging around for lightbulbs?

Yeah, why not just put a damn button on it at that point. I gotta go up there and get the light bulb to pull it down and read the code off of it. So I know it's for lawyers, don't go screw it back in and switch my light three times when we wait, hang on whistle Dixie while I reset it. But

you're. So you have to get the ladder to go change the light bulb. You mean like the way we used to do it?

Yeah, but this reminds me a lot about back in the early 2000s and 90s. When you had to do like reset procedures or diagnostics and cars, there was always a dance you had to do with ignition and

the pedals. And the driver's door open and closed, open closing and

sometimes it was the the lights and stuff. And so people were making fun of this one button that you had to turn on and off in the sequence five times to the sink of a video. Whereas back then you're reading on a form, like turn off and on a light of the key three times and pause in the on position and then back into the accessory position while pumping the accelerator. And it's like you kind of have to like guess the timing on that you didn't have a video showing you how to do it.

See that's why that's why video games got it right back in the day. Because if you hit the

secret stuff down Ab Ab left right left right start select

it would be a sound after if you got it right. It would be like Yep, there you go. Yep, yep, infinite continuous. My car has infinite continues now.

And two player mode as well. Right. With the Select was four.

Oh, geez. We've been going for a while haven't we? Yeah, I

think we should wrap this one up.

Yeah, yeah. This is this is one of the lengthy one Once

Ah, all right, well that was the world's longest Makerfaire fab engineering podcast. I was your guest Josh Rojer

and we're hosts Parker Dolman

and Steven Cregg letter one. Take it easy

Thank you. Yes, you are listener for downloading our show. If you have a cool idea, project or topic, let Steven and Roz and I know Tweet us at McWrap at Longhorn engineer or at I analog E and G, or email us at podcast@macro.com and check out our Slack channel Josh slash Roz hangs out there all the time. If you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest episode right when he releases and please review us wherever you listen to helps the show stay visible and helps new listeners find us