Stephen's finally upgrades his toolkit to include a Digital Logic Analyzer and Parker repairs a Tigershark. Don't worry it's not a real shark.
The Useless Machine contest has just closed. Thank you Mouser for sponsoring our contest and our judges! The Podcast Favorite gets chosen!
Mach 3 Cheeseburgers, Simulations for the PinoTaur mosfet drivers, and more on this episode of the MacroFab Podcast.
Parker
Stephen
R.F.O.
Visit our Public Slack Channel and join the conversation in between episodes!
APA-102C-NEW-260 Test PCB. Arduino shield format.
APA-102C-NEW-260 Test PCB powered up.
DOOM SAO powered up and running face animations.
Back of the PCB for DOOM SAO.
Fixture for DOOM SAO to speed up soldering of the screen to the PCB assembly.
Reading bytes from the “EEPROM” that is the SAMD21 MCU over I2C.
Atmel SWD to Tag Connect adapter. Allows 3.3V power over the Tag Connect.
Atmel ICE programmer connected to the adapter board. Works great!
MacroAmp layout. All that is left is ground routing.
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!
Are you a web that needs PCB assembly look no further than macro fab, we've got you covered whether you need a single prototype or high volume manufacturing, including system integration, macro fab is your solution. Use our easy online interface to upload your files, get a quote and minutes and order your PCB assembly without speaking to anyone. A few days later, your high quality PCBs will arrive in the mail, visit macro fab.com today and sign up to get started.
Welcome to the macro fab engineering podcast we're your hosts Parker, Dolman.
And Steven Gregg.
This is episode 177. Just a quick
reminder to everyone we are well into the macro fab engineering podcast Useless Machine contest sponsored by Mouser. The podcast or sorry, the contest began June 1 and goes through August 10. So just a reminder, it is still open, you can enter in your projects into that. So first place prize is $1,000. Second place is 503rd. Place is 250. And there is also an additional $1,000 prize that you can win by making Parker and me basically left. So how to enter create a useless project, build an electronic project that is completely useless documented on hackaday.io and tag it with macro fab design, contest colon, useless machine, you can find out more about the contest at macro fab.com/blog. And you can also listen to episode 175, where we go into a bit more detail. And we
announced two of our judges so far our guest judges. So we have Joe Grand. And we have Charlene ganda
been guests on the podcast. Yes.
I think there's actually no there's two the two new guests next have not been on the podcast yet. Oh,
that that very that narrowed it down a whole lot.
It's something like 7 billion possible, you know, selections for that.
Yeah. When When are we going to hear about the next judge?
It should be this Saturday or Friday whenever the marketing team decides to push those tweets out.
Got it? Okay, so Joe grinned. And Charlene Gunda this one, I'm really looking forward to this one. It's going to be a ton of fun. I actually wish I could enter. But no, so I came up with another quick idea for a useless machine. And I think given I think I'm I think I'm going to come up with with ideas throughout the whole contest. A fire alarm that only detects when there then only stops when there is a fire. So it's always blaring, it's always blaring and it stops when it tells you there is a fire the fire non detector. Fire non detect, yeah, I like There we go. All right, Parker feed me What's up? What's new?
Okay, so I got those. Those serial smart LEDs tested. These are the APA 102. C, New 260 actually has new in the Part Number,
which means they're all I mean, 260. Yes.
And I tested them. So I built a board that has 10 of these LEDs on it. It's basically an Arduino shield, the poster design files up somewhere. ramp to the reflow. Oven. And they 100% yield.
So did you do one board of 10? Or did you do multiple,
I did one board 10. And so now I'm going to be doing a whole panel. Basically the LED modules that we're using for Penetrator, I want to build a kind of like a small production run probably like 100 of these LEDs, just to see if they, you know is going to be 100%. Again, what's the fallout at 100. And then if that works out pretty well, I'm looking at making these kind of like house parts at macro fab. So now people will actually have an alternative to you know, for these LEDs.
You know that's that's a really this is actually kind of big when it comes to the multicolor LED addressable lighting things. Because there's just been so many problems with manufacturing them, and so many people want to use them, then it's just a giant pain in the ass. So that's pretty cool that somebody's found a solution to it. Yeah, and so I don't have
the APA one oh fours yet. Those are the basically the equivalent to ws 28 twelves, which is what everyone uses. That's like the one wire communication. So the 102 C's have Serial Clock as separate things. Okay, but yeah, when I got the board, the first thing I did I put it underneath the microscope. And so looking for cracks in the lenses, and they all look good, like they're clear. There's no bubbles in them. Nothing like that. Nice. So I was pretty excited. So when I get the one on fours, I'll of course test those. And then we'll push forward to a small production run with the 102 C's new 260.
I'm actually kind of excited because I have right now at work. There's multiple clients that I've talked to who have asked about using well, about using the old parts. And one of them sent me the datasheet for this one. And I was like, I don't know, I'll have to get back to you on that one. So I'm actually going to send them I
should send you some of these. Because I've got a whole reel of them. I can send you some for testing. Cool. You also process?
Yeah, yeah. Excellent. Well, funny thing is, we have the same oven that you do so well.
Different reflow profile. Yeah, sure. So and then, next up is the Doom should he add on? It's actually in production now. Woohoo. So the prototype worked great. The, the SP 7789, LCDs worked great. Like I, you know, had I got a whole box of them from China. And I tested, I basically spot tested, some of them, they all work great. And so we push forward with production, that USB type C is used for the use for USB power and also for communication, because you can do a serial port over it. That works well that works in both orientations. I also got the this was like yesterday, I got the 80 SAM D 21, which is a microcontroller. I got a i square C prom interface working on it. So over the shitty add on connection, there's an ice core C communication for it. And so what you can use that for is there is a unofficial official shitty add on protocol standard for it. So where you basically you put an EEPROM, you can put an EEPROM on your shitty add on that's connected to it. And then you can load in some bytes to it, say, Hey, this is this shitty out on by this person. And that way, the badge, the main badge book, oh, I know what to do with this,
like special embedded code or something
correct. And so I actually implemented that into the microcontroller itself. And so it responds to that a prom i square C address, which is like 50, I think hex 50. Cuz I got that working. I it's not fully coded yet. It basically either works. My code either works is a read only or write only. So I just got to set up a state machine to handle those two states. And I'm using the wiring that like the built in Arduino wiring library. And I don't know, this is one thing I have to figure out is I don't know if the wire library tells you, if the host is giving is asking for a read or read or write. Because I know, you know how many bytes was sent. So if you go, Hey, there's only one byte? You know, it's a read. And single byte read. Right? It's like asking for information. That's the address is that byte, right? And then if you see two bytes, it's the address and what needs to go in that address. So you know, it's a a write operation. Now, I want to be able to support like multiple read rights. Like, like bursts, basically, yeah, consecutive. Yeah, it doesn't support consecutive, but I it supports a single right now. And if I could figure out, I got dig more into the wiring library, if it actually will let you know, hey, the host is saying I want to write a whole bunch of stuff. So I can put it in, I can put that state machine into that mode, so to speak. That's the next thing. It might be I just go it doesn't support it. But it'd be nice to be able to support basically what that E prompt supports directly because then it'd be cool based on how I'm figuring out how to control it is like the first three bytes are reserved for one is like the maker ID who you know who made it. One is the year the DEF CON year, which is DC 27. And then the other one is what should he add on in that maker idea. So if I had multiple like if I had a a Duke Nukem shitty add on. It could do that. First things and then everything after that is just unreserved. And so what I want to do is basically if you write a bite into some of those addresses, it changes what he does on the screen so like his normal is just gonna be him normally looking around but then if you like maybe he has like a damage address so as his damage climbs he gets more bloody Right, right. Yeah, stuff like that.
There needs to be we can have multiple shitty add ons per or is there as always just one you can have multiples. So we we you so it's potential that you could have it where if a Duke Nukem shooty add on and a doom guy shitty add on were attached to the same badge they could fight?
Yeah, cuz actually, they would both respond to the same address.
Right? Exactly. Yeah, that would be awesome. Tara would win Duke Nukem or do or I guess he's not doom guy anymore. He's called Doom Slayer.
Doom Slayer. You're right. Doom guy.
So, actually, real quick I saw on on Twitter. I think you posted it this week. Like, like a 3d printed jig for attaching the screens to the
Oh, yeah. Yeah, sorry. Well, I
remember seeing that because you pray. There was a mistake on the first print. Right? Yeah, it
was half a pitch off on the LCD. So I was like point two millimeters off. Something like that. Okay. Yeah, I think I think the screen is point four millimeter pitch. Basically, I just flooded with my calipers. Right, right. But yeah, I 3d printed a basically a little. I'll put some pictures up in the blog in the blog for this, but it basically holds the PCB in one section of the screen another section and it allows the FFC or the Flexi cable to be in the right spot. So you just have to drag solder across. Okay, so it's sort of like the the base tray of a hot bar solder machine. But yes, without the hot bar? Yes, yeah. And this way it holds it at the right distance apart, because it does have to have the sticky foam and then fold around the PCB in the right spot. So this way, it's all nice and consistent. I
hope. We'll see. Did you do a little cutout in the board for where the flex cable folds over?
No, since the board is the percentage of PCB is actually as tall as the screen? Yeah, there's no spot for that. Got it. Okay. Yeah. Cool. And only reason why the board is wider, is to make room for the breakouts, like, all the pins for the microcontroller are broken out that are not used. So you could hook it up to other things.
Did you use a surface mount header for for the for the actual connection to the to the badge?
Yeah, actually, yeah, I use a two by three. You know, 2.54 millimeter theater. Yeah. Cool. And it seems to work pretty well. Like I actually, I put it into a breadboard and then twisted to see if I could rip that connector off. And it just mangled the connector. Nice. Hey, Damon, rip the pads off. So I'll post those designs up. Basically, what I did is the pads are via stitched through. Okay, board. So like they're anchored to the PCB. Oh, that's cool. And I could not get the connector to rip off. By norm normal means like, I had to take it in like with pliers and like, twist it like 180 to get them the rip off the pad. And even then it actually rip it actually broke the sack through a five solder connection. Didn't even rip the pad off.
Wow, that's impressive, actually. Yeah. So
you could I actually saw her new connector on it and just shipped it to one of my friends who's helping me do the code.
That's great. So is it the is it a standard for the badge to have the male pins?
The badge has the female? Oh, really? Okay. Yeah. Because the males or the male connector is cheaper. And you have more shitty add ons and badgers?
Oh, really? I would have I would have guessed the opposite. The bad would have the male because it's cheaper. And people are trying to cut cost. Interesting.
Now the shitty out is going to be as cheap as possible. Well, yeah,
I mean, that would make sense. Right? But yeah, cool.
Soon this design. I'm going to push it up on a public GitHub repo. So people can start playing around with the code and improving it. I hope because I am not the best programmer. Any any easter eggs. There are going to be some right now. I'm not saying of course.
Well, okay. So the In other words, yes, there are easter eggs there are
going to be easter eggs.
Yeah, cool. Great.
It's probably going to be hunting around in the I square c e prom, you know, structure? Where is it going to? You know, where's it hide net? What address? Is it hide net?
Well, okay, so I guess you can sort of like modern day game, Genie it and pull a look at a bunch of registers, and see if you can find the easter eggs in there. But you know, I digress. The so the fact of the matter that on our notes here says you can pre order it on the and not XR website. There's guaranteed to be easter eggs. If you order anything from there. There's
Exactly, yeah, so if you go to the XR website, there'll be a link on our blog, of course, you can preorder the screen and or the shitty add on and yeah, get it before DEF CON, or I think you can pick them up at DEF CON as well.
How many in total are going to be available.
We are building 200 of them. So that is 200 plus two prototypes.
Nice. So
and then to program them because they have a tag connect on them. I designed a Atmel SWD ICE program or to tag connect adapter. And so we've talked about these before these adaptive boards basically allow us to power and program over one tap Connect cable. And so I tested it, and it works. And on the back of it I put vanilla ice. Because Ice Ice Baby is kind of like a super long running joke between Stephen
classic. So just out of curiosity, have you Did you see this Hackaday article? I think it was what is it? June 13 is when it came out. So not that long ago? For the SLIC byte the soit byte it's a here I'll
Oh yes, yes, yes, yes, yes, I've seen it. It's the little programmer clip.
Yeah, it's a programmer. It's like a It's a programmer clip that basically uses a think of an SLIC style footprint that wraps around the the edge of the board. And it kind of it's like a chip clip. And it clips onto the side of the board. I thought that was a really neat idea. I thought especially as
they're much less expensive than technics tag necks are like, the cables are like 30 bucks. Yeah, those those programmer clips are like, I think I looked them up to like $3 The order of magnitude less. Yeah, the only thing I was looking at is they don't really have least I didn't see, like a positive fixed, like a way like physically will latch on to it. Like I envisioned this thing like sliding and twisting easily off.
Oh, well. So so it is spring loaded. And and so it has it has a row of holes in the PCB, that it loads your fingers grip into the whole thing. And then they touch pads. Yeah. So it probably still can twist and rotate off. But it's probably not as easy as you think.
Well, I'm gonna I, I'm actually going to go in order some just to test them.
I think yeah, I think on eBay, they were like three bucks. Yeah, I
want to test when it's like, hey, is this a valid alternative to Tech Connect? And if it is, then why not use them?
Well, okay, so here's the thing, tag connect, taking their programming header is already fairly small. In fact, they even declare it on their website, that the actual area of the pads is about the same size as an Ohio five chip component, which is true. And so they're already small. But the nice thing about this XYZ bite is that it's on the edge of a board. And it takes up even less room. Cuz because there's there's a bit of even though like the actual pads on a tag connector are small, there's still some keep out you have to do for the holes and clips and things like that. I love them, I think they're great, especially with the fact that if you have a module or something that you have to that has a whole bunch of crap on it, you can access it from the back, you know, a lot of times you don't have access to the side of a board or it's sitting kind of flush to whatever it's mounted to. So that's a little hard for things like this XYZ bite. And in that case, I think Tech Connect wins. But for $3 and a smaller footprint. The SLIC byte is a fairly good idea. Yeah, so I
was gonna give those a shot. Just be you know, I want to try everything And if it works, it's like, Hey, you save, you know $27.
Okay, so we bought, gosh, we have some not Atmel license, what are they called the st link v2, the little, the white beetle junk. So we have some of those that work because we do STM programming. And we recently switched over to using 10 connects because 10 connects kick ass. But if you have the st link v2, in order to use a standard six pin programming header, or cable, you have to have the cable plus the adapter to go into it. The other day, I bought two cables and two adapters, just to you know, give one to the engineering team and one to our production testing team, basically. And the bill was 150 bucks for those four items. He's Yeah, they're proud of their stuff. The Yeah, the cables are $35. But those that a little adapter board, that's just a board and two connectors is like 40 something bucks. That's
why I build my own adaptive boards. Yeah. So yeah, that's, that's what I've been up to. What about you, Steven?
So I finally put the time in and got the macro ramp, rev to up and running, and up and running, meaning 99% of the way there. I have one net left to route on my board and it's ground. And I haven't had hardest one, I went on the I still get this I left the hardest one, but I left an entire layer on the board wide open. So okay, so are you
going to do a so what kind of grounds scheme are you going to do then? Because you have a whole layer? And so for me, I would be like, plain plunge done.
Right? Yeah, I am not going to do that. Well, I mean, I have to I have to do the plunge part, but I'm not going to do a plane part.
So you don't it's a four layer design, then it is a four layer design. So
actually, let me let me back up a little bit. So I'm
gonna back up a little more to is actually the Doom should be on is a four layer shitty add on. Oh, damn. So for I don't know of any other shitty add ons that are four layers? Let me know and, and Twitter, if there are other ones.
All right. And sounds like I'd like to make a four layer. No, I need to make I need to make a six layer one in like a month. Yeah, yeah, that's not going to happen.
Just slap two LEDs on it be
fun. Yeah, two LEDs and six. Actually, one thing I've always wanted to do is, I know this is epic, you know, if I wasn't a guy on the podcast here, I would totally put this in as a useless device. Make either a six or an eight layer board, where you have the minimum space and trace width, and each layer you serpentine a trace such that it consumes every ounce of the board minus the minimal amount. So basically, using six layer board, what's the longest possible trace you could do on say, a one inch by one inch board? And how much resistance would that trace have? And make little tiny heaters? Yeah, make it make a tiny, little tiny heater or just make a PCB resistor, basically. Yeah, these are the things that I think about, there's something wrong with me. Okay, back to the back to the Mac ramp. So I basically took the old design, which frankly, not only is it working like I actually have it built. And it's funny because I built it and I was like yeah, we're gonna do this again. And we're gonna make it better this time. So I basically gutted everything and kept only the power amp which the power amp has five parts in it. And I did it all from scratch again. And the reason I did it this time like completely from scratch is because I actually have test data from these transformers because I actually know what the what the coil resistances in these transformers now and also know what voltages they they produce. So I'm better able to design the power supply because I already have one available. But I also have that you tracer that I designed and made last summer Wow. Was that really last summer? There was last summer Yeah, yeah, that was a while ago. Wow. Okay, so yeah, I got my you tracer. And I went this time from actual tube data. I have a whole slew of tubes here that I got a ton of test points on. And so I'm all while doing all of this. I'm kind of got a write up that's going to end up on our blog about the whole design of this new You tracer. And I stripped things down, I got rid of a whole bunch of extra crap that wasn't necessary in the old one. And in fact, the old design had two new tubes in it. And each new tube has two sections. So it actually has four amplifying elements inside of the preamp. And what I've, what I really came to the conclusion was, at the power amp, I only need 34 volts peak to peak in order to achieve the full maximum output. And a, what's called a new tube can produce a gain of about 11. So divide 34 by 11, you end up with something that doesn't require two new tube elements per print. So instead, I decided to go with one new tube across the entire amp. So each half surfaces, one channel each, and then I made a simple op amp gain stage in front of that. So you kind of get the benefit of the high input impedance of the op amp, and I can fine tune the gain and make everything work out and it still meets the requirement of having a new tube somewhere in it. The One of the downfalls of the first design is that I didn't actually have a volume control on the board. Okay, well, the thing was, like I built this board, assuming that I would have an off board volume controls like a potentiometer of some sort, or you would just control the volume from whatever source you are giving it like, you know, software control, like 30 computer or whatever. Or if you plug your say your phone into it, you just use the phone's volume. And that would be it. So in in the new Mac ramp, I actually have three volume controls now. So I got a little a little tiny wanqi with
it. So how many how many elevens is go to three,
no, get this, this, this is really fun, in fact, so I took a page out of here, I have the book right here, let me read the title of it. Designing high fidelity valve preamps by Merlin,
showing that to me and I was like I have no desire to read that book.
This book is like 500 pages of just like tube preamps valve preamps purely for Hi Fi. But um, gosh, I don't remember the page number. I'm right here. I'm right here. Let me let me one second page 323. If you have this book, you can follow along, he has a hole a volume control scheme for a stereo amplifier that has three rotary switches that are six position each. So the middle rotary switch is a chorus switch that basically controls the volume of both of the channels together. And each one left and right is its own volume control. So you can balance them. Like that. Yeah, it's kind of cool. And the thing about it is, I was gonna I was thinking about using a dual gang pot and having one pot control the volume for both. And that that's okay, but the thing that sucks about it is those pots, do most of the time do not have a spec for the matching both gangs of the of the pot. Okay, so yeah, if you had a dual wiper, yeah, if you have a dual wiper, the wipers aren't toleranced, right, they had the same value. But if you put it, you know, at noon, they're not gonna both be 50%. But with a rotary switch and 1% resistors, I can be pretty close, close. Yeah, yeah. And with this, I can also have it set up where I have specific DB levels. So if you want to set it to a specific like volume, it'll actually be that D in terms of the signal, probably not actual SPL coming from the speakers. But, but still, I don't know, I just thought that was fun. So if you want to control the volume, there's the course not. But if you want to balance the amp, each one, each channel has its own balance, which is kind of cool. That's kind of cool. So and those are, I've moved this whole design to be a single PCB. And I want it to be such that there's no soldering or very minimal soldering to build this thing. So I'm going to have the board made by macro fab. And on kind of the border around where it needs where all the connections are. I did have pads previously. Now there's terminal blocks. So when you connect the transformer in you just slide it in, and you you know, tighten down whatever connection you have. So the output transformers have four different terminal blocks for the the B plus line which is the high voltage, you have an anode connection, you have an ultra linear tap. And then there's one other connection Oh, there's the output speaker ground So this whole thing kind of just drops into the chassis and then instead of soldering things, you just slide them into terminal blocks, which is a lot nicer.
So what pitch are the terminal blocks?
5.08 millimeters?
Okay, cool. I'll because actually I'm going to be up in a minute. I don't know if we can get it built fast enough, but if not, you have to use wire what the wire funerals
for? Oh yeah, yeah, no I'm no I'm totally going to crimp. Yeah those ferals onto the wires and then slide them into the terminals. Yeah. And I picked a specific, you know, the green Phoenix Contact terminal blocks like those classic ones. But I picked them such that they have 400 volt capability which this amp shouldn't go over 320. So, so yeah, all of that's kind of done. What's what's interesting about it is I pared down so much crap in this, that the board is really sparsely populated the and that's basically the board is big, but only because it needs to be wide enough to reach all the components that it connects to
has got big transformers and capacitors on it right?
Well, yeah, and I also switched over to, in fact, I will hold this up to Parker, he can see it on the video, I've got output tube sockets with long legs. So these will bolt right to the chassis, and then the board slides onto it, and then you solder them in place.
Oh, so so no wires, no wires,
no wires to the tube sockets, the tube sockets or PCB mounted, those rotary switches are PCB mounted. So this whole thing should just like pop in as I tried to make this as simple as possible, there's going to be a little bit of chassis manipulation for like a power entry and things like that, because this is mains powered, and you need a fuse and things and I put those off board. So there will be a little bit of modification to the chassis of the enclosure, you could
you should be able to put the high voltage on the PCB. If it's sparsely populated.
Oh, I can but I want to keep that away from those specific areas. Gotcha. So I want the the fuse The power entry and the power switch for this thing to be on the back of the entire amplifier. Gotcha. And it's going to be I need to get either a paddle bit or a forstner bit. And I'm gonna I'm gonna sink them into the wood a little bit, you know, kind of offset? Well classy, almost, yeah, it's a little more classy, you have to like reach into the hole to turn it on, you know, like a little toggle switch in there.
A little gremlin will bite your hand when you reach in there.
That's exactly yeah. So um, as with most of my designs, when I want to do this kind of stuff, I what I usually do is just go into like turbo mode and do like a PCB layout trance, which I did. I did last night, and I just crushed the entire layout. And I was like, I'm done. And I just walked away.
Your eyeballs roll back. And when when they roll back forward, it's 4am.
Yeah, that's honestly I was up till 4am. Last night. I'm running on two hours asleep right now. I think I've done I've had like four pots of coffee. But But no, no, what I usually do is I will just it's sort of like the old method of writing a paper where you just, like, regurgitate or vomit out all of your ideas onto a piece of paper and then walk away and you come back and revise it. So that's where I'm at. I've got the whole layout done, I'm going to come back, look at it, do a review and be like, Oh God, what was I thinking and fix it all. And then and then press
that Trish go over here. Also, I want
to actually throw a quick shout out and say thank you because this project is actually sponsored by Mauser. So Mauser is actually helping Parker and I out with a handful of these projects, because doing all of this is a little bit difficult to self fund sometimes. So thank you very much Mauser, this will be a very cool little project. And I'm gonna have a bunch of write ups on this actually. So there's, I've already started write ups on the design work. So you'll get to see all the tube data that I've gotten the calculations that I went through, I simulated the living hell out of the power supplies and stuff. So I'm gonna build it and see how well all of my designs and my simulations went against it. And you'll be able to read that on the macro fab blog. And when I build the second revision, also have a whole build blog about that too. And it'll probably end up on analog and analog E and g.com. Also cool. Two birds with one stone. On top of that, I think I'm getting the itch to do something non audio related, because I've been just been going ham on that for like the last two years. Yeah, pretty much yeah. I mean, there's there's been some things that I mean, the huge rates are even though it will was the end goal is audio like none of it actually was audio? That is true. I want to I want to get I want to go do something a little different for our next project. So if anyone has some kind of weird ideas, let me know.
What kind of ideas have you been thinking of Steven?
I've been thinking about controllers for my peristaltic pump for my brewing rig. Also, I can't even know you did this. I you know, I don't even know if we've talked about it too much on the podcast. But you designed the octo prover, which was a device that eight separate temperature inputs. Yes. I kind of want to do some some temperature probing of my beer fridge. So maybe something along those lines. I don't know. I want to do something digital. I'm kind of doing desktop prover a decade. Oh, that's right. Hell yeah.
Actually, I mean, one of the things that what are the we talked about them one at one point in time? They're those really cheap temperature probes the DS 180 years or I can't remember what
they are? Oh, yeah, they are the one wire. Yeah,
the one wire temperature probes. I don't know how many of those you can have on a single line. It would be really fun to have like, a temperature Christmas tree, you know, where it was, like 50 of those probably,
I don't think there's cereal? Or honestly, I don't think you can put them on the same line? Because I don't think they have addresses.
I thought I thought you could. I don't I don't remember.
I don't know how you would figure out which ones which then.
That's a good point. Yeah, I don't remember. But, you know, I guess they have they all have their own one data line out, right. So I don't know. Yeah. I'm just coming up with ideas. Or I could just pull one from the old projects list. Although the thing is, I don't want to do something. I'm kind of thinking not audio and all of the things. Oh, yes. Right. Yeah. So let's say let's do something different now. At least at least one different project.
We've got the you should totally. We've got the Raspberry Pi. Classes. Yeah. And you shouldn't make your your tweet your toilet tweeter that tweets the insult toilet. Yeah, the insult toy tweets are rude jokes to celebrities. Right. Right. Yeah. I that class? Yeah. And Scobie toilet humor.
The thing that I worry about that one? I mean, I love that idea. I think that ideas is absolutely phenomenal. The thing that I worry about it is like, that kind of sounds like the sort of project that could go viral. And I don't know if I want to be known as the guy who made that you should totally do that. Like, do you want to be known
as that guy? Oh, yeah, totally.
I mean, I'm laughing. I think it's funny. And so
it'd be great.
Yeah. I mean, that was, honestly that kind of project sounds like, you could rig that up in up on a Saturday, you know?
Well, no, you got to make it so it can survive. The bathroom environment, Steam, heat. Water, it's got to be like IP, it's got to be what?
Fully waterproof. You put it in the bowl. That put
it in the bowl. So you Oh, so you go and see it. Oh, and even better if you can figure out how to charge it off the water flow.
And when what you do is you take these and you you secretly plant them in other people's toilets. So you get this internet of, of insulting toilets that people don't know mesh network mesh network of insulting. Oh my god, that's amazing. Now that's a bigger project.
Start with one. Yeah,
I don't know what we'll think about I kind of want to do I want to make a PCB. I want to make a PCB with a processor on it. I know that super vague, but it's been a while since I've done some firmware. In fact, I would love to do something with an FPGA. I don't know if anyone has any ideas, tweet them at me at analog en je or I guess go to the Slack channel and hit me up. So I don't I
onto the RFO Yeah, rapid fire opinion. So last week, we were talking with Nick with Biddy about there was a small segment in there about the rightness of bananas.
And a lot
like an entire day was lost like entire day of like 15 engineers were lost. In our Slack channel because of this problem, and there's actually a research paper from 1999 on the nondestructive, banana rightness determination using neural network based electronic nose. And it actually has a accuracy of 90.3%. And when they say electronic nose, it's basically a sensor that's sensitive to ethylene, which is the ripening hormone in fruit.
So it just smells fruit for you all day long and tells you what it's good.
Yes. And they have like a chart in there that like what the visual representation of the banana was in the data. It's an amazing paper. So go read it,
we will, why does it need all this like aI neural network stuff? When it just smells something and tells you when it's like passed a threshold?
I'll know because it sounds like you can just do that with an F loop.
It Yeah, right. Exactly.
That's actually the biggest thing. Biggest joke in AI is AI is when you strip off AI is just a ginormous if loop. Right? If chain.
Right? If this if that? Well, isn't that effectively all digital electronics is just big chains.
Yeah, technically, yeah. Right? Yeah, go check that out. It's that that paper is amazing that someone actually did this. And in 1999,
the it's been solved, people go back to work.
And they actually had the part numbers for the sensors. Now I haven't looked up if they're still available.
Banana sensors. Well, we
look it up like voc now, like they were looking at, basically, the reason why they were using AI is because they were aggregating lots of different sensor data. And then whether or not it was ripe or not, and so giving, it's like, it's looking at like five different sensors. And so one could be if one was peeking higher than another one, then it could be ripe still. If and if that other you know, sensor was not in the if statement level yet or at patch the cut off, I should say. So, it's more complicated in that
way. Okay, so ripeness isn't and we're gonna get we're gonna get goofy here. ripeness isn't that like, like, what the find something being ripe? Or a banana being right, like, is there a very specific point at which they're scientifically like, the equation just says equals control? Yeah, like, Well, how do you determine that? Like, it's just the right level of yellow and it has one brown spot, and no, they have a guy that eats them.
He's like, Yeah, that one's good.
Yeah, there we go. No, yeah. No, I got much love bananas. So we got bagel pins and banana controllers are banana sensors, right. Good. Add sensors. Yeah, yeah. Okay, so I wonder what the next food based integrated circuits could be. Yeah. Cool. So in this next RFO, we have sort of a fun little game that park and I came up with right before the podcast. We were kind of like perusing through electronics websites. And we noticed that even like electronics news, I succumb to click Beatty titles. So we found a handful of clickbait style titles, and just put them up here in the in this RFO section. So we're gonna name them. And then we're gonna guess what the article is about. And we haven't looked at any of the articles, we just literally just put down the names that they put out here. So this first title that we found is need more energy storage, just hit print.
So when I read that I thought it read need more energy storage, just hit pint. And I'm like, oh, yeah, you drink beer and gain stored energy storage terms of fat.
Well, yeah, that that totally makes sense. Like that one's not even click Beatty. No, that was just like, need to get fat. Drink beer,
drink beer. So need more energy storage, just hit print. I'm going to guess, like copy pasting an EDA tool, like, for DRAM or something?
Or, or like, like producing battery terminals on an inkjet printer where you can just print the terminal out onto the paper. Oh,
okay. Yeah, that could be cool.
I don't know it's really click Beatty, so it's probably not anything of that sort.
I mean, actually kind of cool. I guess you can make a capacitor On every printer if you could print it down some conductive ink, and then printed a blank paper, blank sheet as your dielectric, and then another sheet of printed like the whole thing, right? Roll that up, bam, right. And actually,
you know, the thing about it is you could print them
each each layer, you print with like a thickening and a thin thinning. In other words, like the dielectric would be wider than the copper plates that you you print down and you can you could stack them and still be able to print on either side, like the actual terminals that you connect to it. So is it possible to inkjet print a capacitor? I think it's so yeah, yeah, that's actually kind of cool, actually. So like, classic inkjet printers have a black cartridge and three color cartridges, right. So that would allow for four separate inks that you could print. So you could you could have at least one of those be a fully conductive print. And at least one of them be a fully non conductive print, such that you just keep going back and forth between conductive and nonconductive. Right. Are you? Yeah, yeah,
you could do that. Yeah, that'd be cool. Oh, wait, that's what that is. I really doubt it. Next one. Next one.
Go hordes call for evidence on impact of digital tech. I actually clicked on this one because I'm like, Lourdes, it's we're in America. And I'm like, What the fuck does this mean? And the thing about it is it doesn't even say where
but like, what does that even mean? Call for the evidence on impact of digital tech. The what?
I'm going to assume lords are very old people. Okay, or they're they are old people. I wouldn't say old people are lords. But I bet you lords are old people. least the majority of them are.
I don't know. There's probably someone listening right now who's like, man, these guys are dumb. You know?
He's a lord.
And he's calling for evidence right now.
Yes. They're going to they're going to like they're going to have the stacks of paperwork. And the the macro engineering podcast being degenerates is going to be like, one of the bad things about digital tech.
Yeah, they the impact? Yeah.
You have a negative impact on society?
Oh, that's great. All right. It's
the House of Lords. So I'm going to bet you England is where that is that we,
we found this one on electronics weekly.
Yeah. But it doesn't say were the House of Lords of because I'm gonna bet you there's multiple House of Lords in the world? I would assume so. Right?
Yeah. I mean, we're probably sounding like, I don't know, Americans right now, not knowing what a lord is. But some kind of have. But thing is like, it's like some kind of government body.
It's some kind of governing body. But it's like, I want to bet you even if we say it's England, because it probably is. There's probably multiple House of Lords in the world. Yeah. And they don't specify which one it is.
Sure. Well, maybe all of them are calling for evidence.
Maybe? Yeah, you're right. All of them. That's why they don't specify.
But they're all doing it from their digital technology.
Yes. I'm reading it on my digital technology.
What about the what about the evidence on impact of analog technology? They don't seem to care about that. Well, not anymore. They used to they Oh, they use Oh, okay. Yeah, they found that rock and roll. They found no no impact. Yeah, that's all it says. Okay, next, next article. It's titled, organic semiconductors, one transistor for all purposes. Yeah. So, okay, so it's pretty easy to read what they're what they're going for, on the title here, one transistor for all purposes. Although, if you've ever spent even five minutes looking at transistors, like that's just a whole bunch of bullshit. Just
I'm writing this a transistor, that's a p type and type. And you just like, flip it upside down or something that works,
ask it to be whatever you want it to be. And it says, Okay,
did you just assume that transistors tight?
I'm a transistor wrist. This is going downhill really fast, real fast.
So what kind of technology? Do you think this one transistor for all purposes, runs on? Is it some kind of new tech or is it
I mean, the only the only thing I can think of is like, like a flawless perfect transistor switch one that is just like it has a it has infinite input impedance. It can pass 100% current with no, no resistance drop. Yeah, no, no drop. It is off at zero volts. And it is on at zero plus D. You know d v like one.
Whatever the baby is over one DVS like one electron,
right? Yeah, it has one electron turns it? Yes. Yeah. And, and what were the superconductor? So it doesn't it doesn't not get hot at all
I'm trying to think of back in. Voltage is potential. Yeah, so yeah, whatever. One electron potential 111 electron volt. Alright.
So yeah, one electron volt. You basically throw an electron at the gate, and it bounces off and turns the transistor on. Yep. Yeah. That sounds like a transistor that could work for most. Cool. There we go.
Yep. So next one is renewable energy won't make Bitcoin green. But tweaking its mining mechanism might sound a little more specific than the last couple ones we've talked about.
Yeah. Okay. So I what I can, what I can guess on this is, they're probably talking about the energy demand of running your, like high end PC? Mining, right? Something like that. So
computational power, run the Bitcoin network?
Yeah. So I don't know, I guess, are they going on? Like, is there a way that you can tell if your bit mining machine is running on pure renewable energy?
Well, I think what this is, is the probably going to tweak the because right now you just have to throw tons of electricity at it to make this network work. This is not even like mining for Bitcoin, it's just like, you still need to do that to like make transactions happen on Bitcoin, because all distributed. And so that takes a ton of power to do right now. I bet you what they're going to try to do is tweak the algorithm or whatever that controls at all, to make it more efficient, but also still, they have to balance, they can't just make it more efficient power wise, because then you just throw more power at it. The but then so that the keep that difficult, what they call the difficulty, which is like how many flops do you have to do to like, make it work? It's really high. So I throw a lot of power at it to make that work. So there's gonna be like, how do they keep the difficulty up while reducing the energy consumption? I don't know.
So, so this is coming to mind right now. So if you think of a processor, or a computer as a whole, in terms of just a box, that power goes in, and it does something, right. The the result of it doing something, everything well, yeah, but but but Right, exactly. That's everything. But in terms of in terms of Bitcoin, think of this. Yeah, the actual Bitcoins that you're mining are entirely invisible. They're virtual, they're just digital storage, whatever. So you dump power into this box, this box, you know, consumes that power in the form of heat, right? Unless it's unless it's in work. Well, right. Yeah. Yeah. So So basically, in order to create this currency, we burn electricity as heat to generate more currency. So is Bitcoin incrementally increasing the temperature of the planet? I guess so. Yeah. Right. Well, I mean, you have to generate that energy to
you don't really so this is a stretching my limits of my understandings of this cryptocurrency kind of stuff. But you don't run, you don't run the mining. Just the mind. I mean, there's probably people that do, but the whole idea is, it's a distributed network of being able to do transactions. And so if I was passing a Bitcoin to you, it's all distributed in we basically the network of all these machines handled the transactions. And now those machines running there. kickback is a get part of the currency by mining it by doing the computational work. So that's how you pay for this distributed network running as they all get a peek the people who are actually running those machines, get a piece of it. But you're right, though is they are increasing entropy in this world?
Well, it's my understanding that mining is maybe I'm not saying this right. But But mining for a this kind of currency is your computer basically making guesses towards a number. And if you're kind of the one, the one magical one that does guess the number correctly in a sense, and I know this is incredibly simplified. But if you're the one that guesses the number, you've effectively mined some currency and it resets and pretty close. Yeah. So at the same time, there's there's effectively like a ledger that everyone agrees on. And the ledger says Parker paid Stephen 500 bucks. And everyone else says, Yeah, I agree. Parker did pay Stephen 500 bucks. And it's pretty close. Yeah, yeah. So all that ledger is distributed. Right? Right, that ledger is distributed across everyone. That way, if one computer says Parker paid Stephen $500,000, but no one else says that everyone can say, well, that's crappy. That's fraud. That didn't actually happen. So you're mining for bitcoins, is both of those things, right? It's yes, you're guessing to try to actually get the currency. And you're also helping man, it's the ledger across the curve. Reverend. I
think that's how it works. Right?
So it's, it's complex, and it's just, but it's also boxes burning energy doing this, you know? Yes.
So I think what they're gonna try to do is, fix that burning lots of energy, but also keep the work high. I don't know how they do that.
Well, I mean, if you do that, you'd be the one of the richest people on Earth, right? That's like solving if
you get the same amount of work out of like, a 10th. The energy,
right? I mean, that is that is actually an interesting thing. Like having having physical currency in your back pocket in your wallet. doesn't burn extra energy. But like having trying to keep this ledger alive constantly does burn energy consistently. Right? So there's like a payment to just make sure that the money stays there.
Oh, you can actually use it to Write Right? Yeah. Interesting. Yep. So next topic. powering devices with a desk lamp question mark.
What do you think a desk lamp.
So this is why matching, you go into your hotel room, and there's a desk lamp with plugs in it and you plug your phone into it.
That's super on the nose.
Because that's like, the only sockets in a hole. Why are those the only sockets in a hotel room? And they're always worn out? Yeah, they're all flop out.
Yeah. Actually, you know, those they look like they're manufactured with like clip in main sockets, you know, like the sockets are not like screwed into anything that just kind
of they just press into the base of the debt lamp. Right, right. Right.
You know, so I bet you this is some kind of wacky article about like a desk lamp that has a solar panel underneath it. And then there's something
it uses perpetual motion of the light the light from the lamp hitting the solar panel, and then powering the lamp and then going back up to the light bulb and back down again.
Well, okay, well there's there's yeah, there's that ridiculousness, but I was thinking more of like the lamp powers a or the the lamp shines on a solar panel, and then the solar panel powers your device, whatever it is.
That sounds really lame.
It's a clickbait article. It's not it's gonna
be a lamp with a, like one of those inductive chargers and debase
the QI chargers. QI X. Yeah, it's funny because I just found out that that's pronounced chi, right before this podcast because we were looking at articles and I was like, huh, go figure? Because I've actually seen a whole bunch of people on YouTube be like, Yeah, I don't know what this Qi you know.
So I actually clicked it and it's really close to your solar panel idea, really. But yes, it is. Basically, using the light scattered about your room, and using special light harvester sensors which sound like mirror a solar panel width like a prism on top, okay, like a dome prism. These that's when we guess that's probably what it is. And that works. So, okay, cool. That's actually interesting. Okay. And then the last clickbait article.
It's beyond one and zero. So engineers boost potential for creating successor to shrinking transistors.
So we're going to You, Stephen? Yeah.
You know, this is not a new and not a new topic.
Yeah, analog computers were like all the rage back in the 70s. I want to say, Yeah, but
they're monstrously bulky and not fast. And pretty much everything. The
next big thing was going to be the analog computer, right? Actually, I'm wrong is even earlier now, it was the 50s and 60s.
So I did actually skim through this article, because I was I enjoyed the clickbait and went into it. So would there I can't remember the name of the researcher, but some transistors have been developed that have his name was Joe, Joe. Okay. These transistors have multiple states in between one and zero and I think it's actually there were two additional states.
So to it, it'd be 1.33 repeating the course and point six six repeating of course, of
course. The Leroy Jenkins straits should just be the LG Oh, that's great. But that's I don't know. That's that's kind of cool. Like I the the question is, does it just have multiple output states or does it respond to multiple input states? I guess it would have to have it would have to respond to multiple input states like how would you produce multiple output states like what triggers a jump from one state to another state
and it was interesting is chose says that this transistor from the outside so we will take your that box that you feed it energy, and it produces heat and work? Yeah. Which is everything in this world? Is if you made the silicone mat with these quad state, I guess, transistors, yeah. A, the computer wouldn't know any better, it would still think that it's a normal binary thing, just because that allows you to shrink let's say an ALU down. So instead of ALU, taking x binary transistors, it takes y which is less than x, quad sisters, quad sisters
there's a quote that you you captured from this great
zinc oxide combined to form a composite nano layer, which is then incorporated with layers of other materials in the Super lattice. And it's just like that that retro and calculator YouTube video. Oh, yeah,
absolutely. It also sounds like even even a little bit the way you said it there sounded like one of the more intense conversations in a Metal Gear Solid game. Yeah. Zinc Oxide combined to form a composite nano layer. What know flocks die. Snake Snake motor logic
transistor was
exactly that. That might Oh, no, you already got a name for the podcast. I just a lot of times while we're while we're going Parker will write the name on there. So I appear up every once a while onto our show notes. And yep, it was already there. Because I was thinking multi layer or multi level transitions would be good. But no, this is going to be episode 177 The Leroy Jenkins transistor
that will get all the clickbait.
That's true. Yeah. How ironic. Yeah.
How many of our listeners know what Leroy Jenkins is though, because that is old enough. That's actually a really old. They weren't even called memes back then. No, no, they were Weren't they? Oh, plea were not called means back then. Because it's like 2007 That's how old that is.
Okay. Leroy Jenkins has a Wikipedia page. That's impressive. It is isn't it? Let's see here. A
meme
may 2000.
Oh my god seven or older than I thought it's oh five. Yeah, May 2005. Wow. Right. Wow. Okay, great. Because that move that I say movie because YouTube didn't even existed.
Known YouTube didn't exist for
four years. Now. YouTube came out in Oh, seven.
I thought it came out nine. No, was it seven?
Maybe you're right. I can't remember when YouTube existed. Founded in oh, five. Oh, I think it was oh six when it came out. Got it.
Okay, well, it's still. I was in high school when Leroy Jenkins happened. Yes, yikes.
I was playing World of Warcraft when that happened.
So you probably loved it. So actually,
oh yeah, that was awesome.
I get how many of our users had no clue what Leroy Jenkins was and figured it out because of this book. You're welcome, by the way.
I wonder what the highest resolution of Leroy Jenkins you can get is? Oh, most of them are awful. It's it's probably like 240 P pro today. Yeah. Yeah. heavily compressed, recorded on a cell phone.
Yeah. That's awesome. Wow, look at all of these references from the Wikipedia page. There's so many references. There's the at least 24 of them.
It's probably more referenced in a lot of other topics on on Wikipedia.
There's a whole like, yeah, here. Here's a whole YouTube video of what is Leroy Jenkins? It's an explanation for those who don't so great. Yeah, the good that that honestly was, that was like, the spawn of means right.
It was one of the well we'll get one of the first themes is like, Well, what we would call me is like the dancing baby GIF.
Yeah. Well from Ally McBeal. That show or that show back way back in the day, and like that was back when II bombs world and all that stuff was like,
a thing. You know, it was the only thing out there Well, there
were there was FARC you remember? fark.com? No, I don't actually FARC no bombs. FARC was like E bombs. It was just an aggregator of weird stuff. Well, I mean, it's tame by today's standards. But back then that was the stuff that was like Yeah, you were you were like underground if you want to eat bombs. Ah, I think that's probably good for today, don't you think?
Yep. So that was the macro fab engineering podcast. We're your host Parker Dolman and Steven Gregg. And at least I had chicken.
Leroy Jenkins.
Let everyone take it easy. Thank you. Yes, you are listener for download in our show. If you have a cool idea, project or topic or meme. Let Stephen I know Tweet us at McWrap at Longhorn engineer with no O's, or analog E and G or emails that podcast at Max web.com. Also check out our Slack channel. If you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest episode right when it releases and please review us wherever you listen as a helps the show stay visible and helps new listeners find us. Please review us. Please. I read all our reviews
Stephen's finally upgrades his toolkit to include a Digital Logic Analyzer and Parker repairs a Tigershark. Don't worry it's not a real shark.
The Useless Machine contest has just closed. Thank you Mouser for sponsoring our contest and our judges! The Podcast Favorite gets chosen!
Mach 3 Cheeseburgers, Simulations for the PinoTaur mosfet drivers, and more on this episode of the MacroFab Podcast.