Failing to Electrocute the Chicken

Hello, and welcome to the macro fab engineering podcast. We're your hosts, Parker,

Dolman and Steven Gregg.

This is episode 241. So we got some feedback about our idea to live stream a podcast on Twitch. I think people want us to do it on YouTube Live. So we could do that it's just different platform.

It might be more accessible to a wider community.

Yeah. So we'll set that up. And we'll do 242 for that. So that'd be six o'clock, estimated time, Tuesday next week at six o'clock central time. Right? PM, not at 9am No one wakes up that early.

Not us to do a podcast that's for damn sure.

So I want to talk about one weird interesting, actually, oh, my topics kind of weird and interesting for this episode. But um, so this weekend, I was working on the wagon, prepping it for a larger radiator, and cooling system. And realizing it's all the little things that you do on a project that take forever and no one sees.

Oh, yeah, they see you start the project. And then they see the final product. Yeah, but

like, so I had to move the radio, the AC condenser Ford and in the in the, in the wagon, so that can have more clearance for the radiator. Because I'm putting a bigger radiator in and and just the move the AC condenser two inches forward, I had to reroute all the transmission lines, reroute the power steering lines, move the coolers, massage the metal a little bit so that it fit the brackets better support costs of maintenance, yes, and progressive maintenance with a with a rubber mallet. And then, you know, remount, the AC condenser and had a you know, wait for my friend to show up so that he could hold it in one spot because it was like in the engine bay. And you had to drill through the the material to do that. And it's just like, No one isn't going to took that was basically about six hours of work. No one is going to see this at all. All that work. No one's going to see it.

But you'll know that you did it.

Yeah, exactly. But it just reminds me just stuff like that, that just takes forever for a project. And it's it's the stuff that people never see.

You know, an interesting parallel that goes along with that is when when you're asked to do a revision on a project in a in a more professional setting, and there and you're asked to just just change this one little thing. And when it comes down to like, okay, now I got to look at my schematic, and that one change causes a huge ripple effect throughout the schematic, when you're talking about like, okay, my net names are all going to change. So yes, it's it's one thing, but like 50 nets now change name or something of that for it, or like, Oh, I've got an inversion incorrect, I need to add a half of an op amp and change that the ripple effects of what that does to your layout. And like, Oh, my God, like my layout was already tight. Now I'm gonna have to spend forever just inserting another op amp stage or something like you. Those are the things that people don't pay attention to

it, there's that. And if you've made a major change to your part libraries, like about six years ago, I did a major change to my part, my Eagle part libraries, I just went through and redesigned pretty much every single part of my libraries, my personal libraries, just to bring them all into like the same standard of like, how they look and that kind of stuff. And then I open up one of my old PCBs and hit update library. And everything gets messed up everything just bloated, right? Yeah, everything's exploded. So you're like, Well, time to go back and redesign this whole board then?

Yeah. I think I think if you are a manager, and you are somebody who is passing down work to engineers, or anyone who's working for you, like it's it's a very valuable trait to understand the magnitude of what you're asking for. Because, yes, the the concept of like, I want somebody to make this one little change might result in something that is far greater than that one little change. To accomplish that one little change. And just keeping a good I don't know. I've been reading a book recently that has some concepts about

mindfulness is a word that they use in there and it's like a trait of people who are successful in a way and This isn't one of those books that's like 10 things that successful people exhibit or whatever. This is just a. Well, the concept is mindfulness is being cognizant of these kinds of things, as opposed to just like, I want this thing done, go do this thing, regardless of the consequences, you know?

Yeah. It's like was the butterfly effect? Right?

I think that yeah, sure. Changes 10805 10

Oh, 603. And like a volcano explodes on the other side of the planet.

There's probably actually some truth behind it. Cool.

Cool. What are you been up to Stephen? Okay. So

I've got I've got a little bit more of an update on my high voltage, low current measurement. I my topic thing is the circuit evolves. Because I don't want to say feature creep. I want to say it's evolving. Expanding? No, no. Okay. So so I shouldn't say this is feature creep. You know,

I just came up with a really stupid idea. What, please share it, well, maybe not idea is the right word. But concept. So we have feature creep, right? Yeah. What if it was your your business so stupid now is done thinking about it is like, you know how like, your PCB is like solid. If your PCB was liquid, and it could expand to the volume of the container that is your project. Or it turns into a gas and further expands.

I'm gonna ignore this Akiko.

There's gonna be one person out there, that's like Parker, I get you,

I get you on that. So actually, funny tangent on that I took a class. I don't even remember the name of the class. It was it was a higher level class in college, but it was we thin film technology. That's what it was. It was thin film technology. And we talked about, funnily enough, we talked about sunscreen a lot in thin film technology, because the professor was really, really, really into the concept of using sunscreen as an analogy for electronics. Because it's like, it's a really thin coating you can put on your skin, and then it does something. I know, Parker's face is golden right now, because he's, you know, okay, well, I can see where he's coming from. Sure, sure. And it wasn't she or she,

my bad.

So the thing is, Okay, shoot, where was I going with this?

It puts the lotion on the resistor or it gets the edge again.

Okay, so it says the funny thing was, at the, at the, at the end of one class, I went and talked to the professor about an idea I had, and she was like, Do you want to go to grad school and do this idea? And I was like, No, not really, I just want to talk to you about this idea. But but one of the concepts was, could you create a paint that the paint itself had p and n material in it, such that when you mix it together, you get P N? LED junctions? And who cares? How many are bad? How many would be good? Is all that matters is like you mix this together? And then could you somehow apply a potential to it, and, you know, basically paint something that can emit light. So you could effectively illuminate your walls or illuminate your ceiling, like have the entire ceiling like slightly dimly glow? And then you don't need like light bulbs or anything, you just illuminate the paint?

So you would have to make a you'd have your paint would have to have you'd have to dope the paint. Yeah,

effectively. And there's, there's, there's like a bazillion reasons why that wouldn't work and why it's incredibly difficult. And how do you apply potential across a liquid, like throughout the liquid, you know, but your your idea of the like, expanding the liquid PCB reminded me of that idea. It would be super cool, especially if you could somehow do it, RGB. And then like, just control the color of the paint on your walls in your house. You know, you like I don't like white anymore. I want a pink house. And then you just get on your app and make it pink.

I like that idea a lot. Yeah, it's crazy. I think the name to dope paint would be pretty good until it's paint that's doped with different

elements. Yeah, yeah, that would be Yep. Right. Exactly.

Just don't look at the ingredient list.

And the whole concept is like, you have P stuff and you have n stuff and you paint it on a wall and it dries into the zillions of diodes that are facing in all different directions. And, and then somehow it just illuminates light, you know, on directionally Yeah. She was the professor was like, let's do that go to grad school with me. I was so done with school by that tower. Yeah. Yeah, I think I was a senior. It was either my second to last or my last semester. I was like,

Yeah, it's interesting to how many, at least in electrical engineering, how many students when you were a sophomore, everyone's like, I want to go get my PhD and you get us be a senior. And it was like, fuck this shit.

Yeah, yeah, it was it was gotten to like that. My last semester I did, I actually did really well. I think my last semester I even got on the Dean's list. But like, I remember that entire semester was just like, trying so hard not to clock out.

I really liked my last semester, right? My senior year was my favorite year.

My junior year was awesome. That's when I took all the best classes.

My junior year was also really good. It just, it was like a just peak senior year, I was just like, This is awesome. And then the worst thing though, about it was when I graduated, because we went to school during the economic downturn. What Oh, 811, basically. And so at UT, they had to cancel a lot of classes, because they just didn't have funding for them. But they started to bring back all those really cool like robotics and like antenna classes. And I'm like, Man, I wish I had one more semester just to go to those classes. Yeah, that's a good sign, though. It's like when you finally get to the anyone that's in electrical engineering right now. The first couple years is a slog, because you're just doing the fundamental classes. And that kind of stuff, though. It's grunt work, the matte matte really math heavy stuff. Yeah. It's math for math sake, as well. Yeah. But when you finally get into your junior and senior classes, you start to actually do application stuff, you're still doing math, but you see a real tangible thing with that your math does. So it makes it worth it.

Right. And eventually, it's funny, because like, in the in the calculus classes, and even some of the physics classes, you learn a bunch of identities and you learn, like trigonometric stuff, that is just unbelievably annoying. And then it actually shows up later, and you have to use it. And it's like, okay, this is good. Yeah. And if EQ starts to make a little bit more sense later on, I

think I got an aneurysm when you said defeat, you know, I remember taking a, a class about it, it was, it was filters, but it was just math. Filtering, just met the math behind filters. And it was like, a sophomore level class, and I'm like, This is impossible. And then but it wasn't until like my senior year, I took a DSP class, and then actually having to do that math again, and actually implemented though, and software and run some basically built a modem. And then I was like, Okay, why would why wasn't like this part of that class? You know, because half of my friends in that that filter class, the math filter class, like just dropped out. Because it's just like, I don't see the point of this, you know,

edit class similar to that. And frankly, I love that class. loved I loved doing the Laplace transform and finding transfer functions of like, complex filter circuitry, but I don't know just I, that was fun.

Well, maybe you saw what the end goal that was already. Oh, yeah. Yeah, me being digital boy. Didn't see that.

Honestly. I would. That was that was exactly the same for me when it came to my first digital logic class. Like it was so poorly presented, that I didn't, I couldn't grasp what they were getting at. Like, I was like, okay, cool. You taught me this thing. What's the end goal here? And like, there was never an answer to the end goal. It was just like, draw this Karnaugh map because that reason word I'm like thinking of Karnaugh map. Hey, maps are the worst thing known to man. Like there are so so good. Accardo maps, I hate Karnaugh maps. And like, I've the only time I've ever used heard of or seen a Karnaugh map is in that class, ever. I mean, don't get me wrong. There's Karnaugh maps on like data sheets. I get it. Yeah, but

I've used them in a while develop. Well, back when I did a lot more FPGA stuff. I would use Karnaugh maps a lot. Yeah, screw contour maps. But yeah, there's not a lot of use cases for them anymore. Mainly because like Verilog exists now. also that you could write readable code that compiles into hardware for you already. So you don't really have to, like optimize a bunch of and and or NOR gates like.

Yeah, right. I don't know. There's someone, there's someone out there who probably really loves them and more power to you.

Yep, I get you.

I'm gonna add to this like, tangent trail here. But I probably said this before, but the digital logic professor at a&m was, like, renowned for being weird. And his, his description of logic gates was electrocuting chickens. Like, I think you told us Yeah. And it's funny, because like, somebody talked to him about that and be like, you can't, you can't really do that anymore. Like, that's not a good way to teach. So he changed it to shocking snakes, and they'd like whip their tail when you shocked them. But if you don't tuck them, then they don't want their tail. That's how he explained logic gates, I swear to God, like I paid this guy $300 an hour to tell me to electrocute chickens. And somehow that represents a logic gate. I'm a little pissed off about it still. That's weird.

How does that have that work? Like that example work? Like, like, take, because you like, I'll put this way you apply enough energy to anything it's gonna wiggle.

i He was he was literally just trying to Yeah, he was basically just trying to show that like, if you do something, do something, you get a result. I mean, I don't know, I've erased that whole class out of my memory. It was awful.

I there was however, there was one cool thing, I failed this on the exam. And I don't know how it failed. Like, I didn't get any feedback on my exams, I literally just got, here's your grade. That's it. And I just I don't know what I what I got wrong. And I would love to know what I got wrong. Because there was there was one, there was one thing where we basically had a processor design on a LM exam. And there was 20 lines written next to the exam, or right next to the image of the processor. And the the question said in in 20, or fewer commands, make this processor do a thing, I don't remember what it was, it might have been like, add two numbers or something. And there was like memory, and there was the ALU. And there was working registers and things like that. And I loved that, because that was super, like, I'd seen that before. And I was like, this looks like the internal on optic microcontroller that I've messed with before. And, and for some reason, I got that problem wrong. But I mean, I made it add two numbers in less than 20 cycles. But I probably, I probably failed to say at the end like you have to electrocute the chicken in order processor work.

I don't know. Good, good stuff. Okay. So, so reversing this, like 20 minute tangent here. My high voltage, low current measurement circuit. So here's, here's what I've what I've done in terms of the evolving nature of it. So I had the whole circuit done, but I was thinking. So the the,

and this is the this is the circuit to measure the current in a tube, vacuum tube,

yeah, multiple currents actually. But yes, to measure to measuring high current, low currents in high voltage situation situations. And I'm my use case is vacuum tubes in a guitar amp. But I've evolved it to be more of like a user function Testbed as opposed to just like a widget that I have laying around. And so one of the things is, I mentioned last week, or the week before, that, in order to accomplish the measurement, I'm using the method of measuring voltage on two sides of a resistor, and then using the resistance value to calculate my current. Well, that means that you have to know the resistance of there. And of course, like, I know what I've populated there, but if I want higher accuracy, I should actually measure the resistor. So that's what I'm adding to the circuit is a small extra little circuit that I can measure the value of each resistor, store that in memory, and then use that for the calculation. And the way I've determined I'm just going to do it is fairly simple. I in my circuit, I have a five volt reference. So it's, it's, I don't remember the absolute tolerance of it, but it's a pretty high tolerance. In fact, somewhere around here, yeah, right here I've got my little five volt reference PCB that has the reference IC, which is an AED 584 on it, and that spits out five volts directly and it can force some current. So what I'm going to do is have two jacks on the front and my box that you can plug multimeter plugs into and it just spits out five volts off of one and then there will be a reference resistor inside of my, my box. And basically whatever, whatever voltage is developed across that reference resistor, I can, I can determine what the resistor I'm actually measuring with. So little bit of adjustment to the circuit, but actually pretty minimal impact, I'm adding another A to D converter on there and then a reference resistor. So here's the process that I'm thinking that the whole box is going to go. There will be a like a computer UI that kind of steps a user through, you know, the whole process of things. But effectively, what you do is with the guitar amp turned off entirely, you use the reference resistor, measurement probes, or and you measure each one of the resistors in there, so there will be a screen resistor for every tube, and then there will be a cathode resistor for every two. And the screen resistor is going to be somewhere in the range of about 100 ohms to maybe two kilo ohms, somewhere in that range. So you'll read each one of those, and it'll store in the in the user interface that value. And then you read all the cathode resistors, and you do and then you store all those also. So then you take my little wire harness that I have that has like a universal connection on it, you plug it into the amp, turn the amp on, make sure that all of your controls on the amp are zero, you don't want any dynamic, anything that we're just reading DC stuff, let it warm up. And it'll automatically read all of those, all the all the currents in there. And then it'll tell you how to adjust the bias trim pot to make sure that the current is correct in each tube. So it's pretty straightforward. One of the things that this solves this whole test system, because this is kind of I mentioned it before in a previous podcast, you can do all of this with a crappy multimeter, like a $10 multimeter. Like, why go through all the effort to do all this? Well, one of the things that this solves is that most amps of this sort do not have a regulated power supply, they are typically running just a linear power supply. So the load on the power supply will change the voltage of the power supply. Because it it's semi regulated. But what's changed is the current which changes the current. So as you turn the bias pot to allow more current India tubes to heat them up the load or the actual voltage off, the transformer will drop, which means you need to add more current, which means that it'll drop further. So eventually there's this magic equilibrium point that if you know what you're doing, you could test all of these points and figure that out. But it's iterative with a multimeter. I don't want it to be iterative, I want it to be dynamically shown on a screen and all you have to do is plug in this test harness Well, after you've measured these resistors you turn the bias pot and it's actively measuring all of those values and it gives you like spot on exactly what you're looking for. Well, understanding that things are the load is dynamic on there. So I'm solving something that doesn't need to be solved. But like, damn it, my stuff's going to be accurate. So yeah, that's, that's pretty fun. So basically what I have left on this, um, I've got most of the stuff, I've got the virtually the whole circuit in there, I need to add this resistor measurement circuit, and then lay out the board. And frankly, the board layout, like I'm looking at it, it's like, I'm thinking this was like an hour or two of work, it's very minimal. And I don't even think I'm gonna go through all the trouble that I've done in the past of like, 3d designing the enclosure and all that stuff, I'm just going to make the board because frankly, this board has three connections, it has a USB connection, which the USB is going to supply the entire power to everything in there. It has the wire harness, which is 16 wires, and then it has the bias probes, which is two wires. I don't mind putting a grommet in the side of an enclosure and putting 16 wires through that grommet like this doesn't need to look pretty, like I'm totally fine with that. So a whole for USB or maybe even a USB pass through in an enclosure. And then the 16 wire note and the two wires

no you tracer style box. Ah,

yeah, I think I might go with that. It's expensive, but I mean, this whole project is expensive. Like I'm not trying to cut corners on this. I mean like a lot of my chips are like four or $5 apiece and those and the resistive divider ICs and organize these those big resistor like sale resistors are like 16 bucks apiece kind of thing. Yeah.

Resistor ladders, right,

right. Yeah. So um, yeah, I am thinking about using that. That lovely little Hammond box with the wooden ends because you know if you're going They go through the trouble of making a test box it should look kind of sexy. Right?

Exactly. That's where that's where we all strive for

sexy tech desk boxes. Yeah.

So the was a while ago talked about upgrading my 3d printer. I finally got all the parts in for that. So I put new fans in it upgraded the heated bed. I say upgraded the heated bed I love to heater the same, I upgraded the build surface on it. So it's running P e i build surface, which is just a fancy plastic material that a polycarbonate likes to stick to. And it's a it's a magnetic base too. So like the the, the metal pops off, so you can print on it. And pull the when you're done. Pull it off and you can use flex it and snap the parts off. It makes it really convenient. And now it doesn't make it well. It still makes noise it makes you know stepper motor noises but the fans are not all crunchy anymore. Which is nice. Yeah, I actually, I'll I'll show you after the after the podcast, but like one of the fans like it's Nachi it's like that's not supposed to be like that. Tasty? Yeah. Yeah. Been printing again. So that's not that's good.

What's the what's the first thing you printed after adjusting everything?

20 millimeter calibration cube.

Was it spot on?

It was it was pretty close. The X and Y was really good. Z was off by like point two millimeters. And for me, that's plenty good. It's like I don't really care too much about point two millimeters.

It's a little less than 10 thousandths of an inch. Yeah,

yeah, it's not too bad. I could probably dial out, it's mostly all in the first couple layers. Because that's polycarbonate, you kind of need a smoosh the material into the bed. So it really sticks. So you probably this, I can probably actually back off that a little bit. And I'll probably bring that point too to only be in point one off or less.

I remember you sent me a while back an article about a guy who did like a Z probing thing for for correcting the table.

That's what I run.

So you actually have that? Yeah, cool. What is your actual like? Like, how what are you doing for the actual probe itself?

So I use a probe called the BL touch. I call it the BLT Ouch. But whatever. I think it's the BL touch. But yeah, it's it's a mechanical probe. And so it will extend down it has a cool solenoid in it. And then look, I think it's a little catch, basically. And so when it touches something that retracts, it flicks back up and sets off the sensor. And so I basically probes nine times around the bed, and then builds a mesh of what the bed looks like. Because it's it's flat in quotes, aluminum, which of course, when you heat it up, it starts to warp a little bit.

Well, and that's exactly what I was. I was thinking like, it's, if yours is magnetically held, and you can pull it off every time you put it on. I mean, if you're trying to hit sub point one millimeter accuracy on on your Z height, like you're not going to get that unless you have some kind of calibration, you know? Yeah, so

it does have a calibration routine every single time you start a print. Yeah, because it will change because as you said, the aluminum flexes a little bit and that steel sheet you can pull on and off so it's not fixed. And so that flexes a little bit too. But um, yeah, I'm liking it. The new heated bed setup, the magnetic base is pretty nice. To see what the prints like I come up with.

I got an interesting calibration conversation real quick. The other the other day, I think it was last week, maybe the week before now it was a week before I was replacing the spindle and our CNC at work, which we have a pretty nice spin spindle there. So it's a pretty nice CNC. So replacing the spindle is is kind of critical getting it like spot on. Is that spindle under warranty? No, no, it is not. No. I mean, the warranty cost is as much as a new spindle. So it's just like, I mean, when we didn't buy a new spindle we bought a refurb spindle which is still rated for a long time. So the thing is with with this machine, it holds such ridiculous tolerances that replacing the spindle like you have to check and make sure that you're in on those tolerances. And it sucks, I spent two days calibrating this machine. Because, you know, at some point, like, the computer is not going to do it for you like, at some point it won't. You gotta get in there and turn screws and feeler gauges involved, there could have been the end and I'm super lucky, I was able to get it in trim, without feeler gauges. And it's funny because they provide a pack of feeler gauges, and each gauge, they call them foils. And each foil is, it's, it's 16 microns. Like, that's how fine like they like, they're like, be really careful, because if you'd like touch it, the wrong way will just fall apart that you know that this thing. Yeah, so So here's how it's kind of strange. But but here's the way that I got it to trim. So I put down a piece of aluminum and I just milled a circle in the top of the aluminum that's, you know, 10 thousandths of an inch deep, and maybe three millimeters wide. So it's just a flat circle. So I now have a flat reference plane that my spindle, you know, milled. So I chuck up into the into the spindle along arm that on the end of the arm, I have a dial indicator on it. So if I'm sitting straight above the, the circle that I milled, I can move to the left or to the right in my x direction, the length of that arm, such that the end of the dial indicator is touching the circle I just milled and I can zero out the spindle or the dial indicator by moving my Z up and down. And then what you do is you move the spindle, the the entirely opposite direction, the length of that same the length of the arm that you're measuring, such that you're basically spanning a circle whose radius is the length of the arm that you have. So the length of the arm that I had chucked up in there was 220 millimeters. And so the diameter that I could span to check the trim of the device was 440 millimeters. And the whole goal is if you're on the left side of the 440 millimeter circle and you zero out and then you go to the right side of it, and you go to the same zero, you want your dial indicator to be identical. That would mean that the circle itself is not tipped, and you're perfectly square to your bed. Well, the first time I installed the spindle, I was like 2000 off, which, okay 2000s But like, that's nowhere near good enough for this machine. Like, that's awful for this machine. And after playing with it for two days, I got it to nine microns across 440 millimeters. So I mean, that is ridiculously flat. But talk about a frickin headache, because you got to take off all the brackets, kind of just wiggle the spindle a little bit to you know, tighten everything up mill a new sphere or a circle, and then do that whole test just to see if you got it. You know, like and the

thing is, if you take it all apart, you couldn't make it worse. Oh, yeah. Oh, yeah,

like and I was on I was on the phone with the service department of the CNC. And they were like, look, we know this is super frustrating. There's just there's not, there's not like set screws that you can like, yeah, angle the ring with you literally just undo the bolts, and then do the back to back in. And let's just have to get lucky that it's going to line up and I didn't get lucky for a while. And eventually I once I hit nine micron, I was like I'm done. I'm done. I'm not getting any better than nine micron. So actually, I faced a piece of aluminum that was 12 inches by six inches the other day with an eight millimeter bit. So and you know, you have to take a handful of paths because I was only doing half of half the diameter pass each time. So each swipe across it was four millimeters, four millimeters, and you can run your finger across it and just barely feel the difference. So it's not perfectly in trim. And I wouldn't ever expect it to be with you know, the way that I was doing it. But like nine micron was like it's almost a mirror finish.

Yeah, yeah. Cool. Yeah, we just pay someone to come through our scene, our pick and place and stuff.

Yeah, yeah. Well, yeah.

Yeah. But yeah, they come in for Micronics. They come in once a year to recalibrate everything. Yeah. And but they have a whole kit. Like you said, you were surfacing the thing. They have a kit that's like that, that adds a plate is what they call it. They put it in the machine. They spend a whole day tuning up all the machines.

You know, it'd be interesting thing to know, like, do they have to do much to the machine? I mean, I'm sure most of it's them just checking if it's okay.

Oh, no, because I've never, I've only watched them do it for like 30 minutes. And I'm like, okay, that's boring.

I mean, the last time we had the service tech come in and check our CNC, the guy didn't really do much. There was it was ever so slightly out of square. And he did just a very minor amount of adjustment. But he spent most of the day just like, he checked it and was like, Okay, this is good. And this is good. And so yeah, we got a test report that shows that we're good. And you know, that gives you a bunch of warm and fuzzy feelings, but it was like, Oh, they haven't you spent all day here. It didn't have to change anything. I guess that's what you really want, though.

Yeah, that's what you really want. And the main reason we do it is because it keeps our machines in warranty. Right, right. Yeah, that's the main reason to do it. But um, yeah, they get you. That's a ligature. But I mean, I don't know if they've ever I had to actually tweak anything in those machines, because they run pretty well. Pretty well. I mean, we run them. Like, what, 20 hours a day or something like that. So

yeah. In the last 10 days, I've that seems he's probably been down for 15 hours.

Is it running right now? Oh, yeah. Oh, yeah. No, most most

of the time, I tried to set that machine sets that I started, like, around noon, and and it ends the program at like, seven or eight in the morning. So I have a little window between when I get to work and lunchtime where I can set it up for its next run. And I try to make all my programs around 20 hours if I can. It's pretty good. Pretty good idea. Yeah. I have to touch the machine once a day. Yep. Unless it breaks and then I have to touch it constantly for a week and a half.

So got another topic here is it's it's a article that found on fierce electronics called ti calls for lower cost sensors to boost robot adoption. And I was reading this one at lunch today and it's it's kind of a weird article because it's ti telling other companies to make stuff cheaper so they can sell more product is what this article is. And this comment here we don't see as many robots as we'd like. And then like you would put in like quotes or in like brackets there so we could sell more stuff.

Oh, for sure. Absolutely. Why are there not more robots? We have all the sensor technologies.

Yeah. But they're saying stuff like expensive modules like sensors and cameras are why we don't have as many robot arms as they'd like to see.

They would like to see them on every street corner yes directing traffic

but yeah, there's some very interesting very interesting what's a good word for it? Examples I guess the the Wow, who is this person?

The Vice President of processors at Texas Instruments.

Ah, yes. Campaign nella? The how you pronounce that?

Where are you seeing this?

That quote I just said we don't see as many robots. Oh,

yeah. Now let's jump in now. Giovanni Campanella. Oh, he's a systems manager for robotics at TI. I was reading a quote from another guy.

So Ti, why don't you make all these expensive sensors and modules cheaper.

Okay, so a little bit back on my, my high voltage low current measurement thing. I got another quick little topic that was falling from that. So I said earlier that I was wanting to do some multimeter probe connections into the enclosure for that. Well, I started going down a path of wanting to put banana plugs or banana Jacks I should say on the side of the enclosure, but I wanted to get those multimeter banana Jacks Do you know the ones I'm talking about? Like they're the banana jacks that have that jacket? That also Oh shielding on it the shields that will plastic shield thing right? Yeah. And so it was like okay, this should be pretty easy to find right? The show notes

because you have high voltage right? It's

because I have high voltage but also I just want it because if you want to if you have those kinds of multimeter plugs, then you you have to have that to put plug them in right You know, the the multimeter plugs that have the plastic sheath on oh, Banana. Yeah. Also the the resistor measurement isn't going to be in a high voltage environment because the amp is off. So got that doesn't actually matter. But still I want it because I like those and they look nice and they feel nice. And they kind of also hold your, your multimeter plugs captive in a way there's they're a little bit of a friction fit. So I started looking for those and I realized like, I don't know what those are called. Like, what? I know what a banana Jack is. And I know what

did you did you type in JST connector?

Yeah, and there's 1000 of them, and they look nothing like it. Right? Yeah. That would be funny. Yeah, I think we should just call every connector a JST connector, like just every single one. Okay. So, and okay, so the funny thing is, like, where I'm going with this is I started going down a path of like, I have, actually, let me ask you, How many times have you had a connector in your mind? You know, it exists? You could, you could picture it in your head, you just don't know the name.

I mean, I could taste that connector.

Exactly like, and you could probably already think that there's like panel mount versions, there's probably PC mount versions. I've seen the inside of a fluke. And they have custom versions of them. Like, they're just like, they're ubiquitous, but you don't know the name. So it's sort of like, I don't know how to spell a word. So how can I look it up in a dictionary? Right? I know what the word is. I know what my connector is. But how do I find it? And technically, I'm looking for a binding post. And after a bunch of searching, maybe it shouldn't have been a bunch. Maybe I should have found it easier. But they're actually just called safety binding posts, which, okay, fine, easy, easy enough there. So I'm going to be trying to find a version of a safety binding posts that'll work with this. The funny thing is, in all of this searching, I ended up on a Wikipedia page for banana connectors because, of course, there's a Wikipedia page for banana connectors.

And then there's one for JST connectors, which is a list of all connect

you know what we should get? Oh my god, we should get T shirts made. That's just the like the black like outline of a JST connector and underneath it, it just says connector, you know. But, but Okay, so the funny thing is, so Wikipedia has a whole story about the history of the banana connector, and apparently, in let me learn a Sunday connector. So there's a way I got it pulled up here. In 1924. Two companies said that they designed the banana connector. One was the Hirshman. What is the Hirshman company? Yeah. And they claim it was invented by Richard Hirshman in 1924. And then a competing claim was made by the general radio company, which stated in 1924, that they had created the banana connector. Although what's really interesting is in their claim, they say they created it, but they said they introduced in this country, the banana connector, which that seems a little bit shady if you ask me. But it's funny because the Wikipedia page actually has a whole article from many years later. 1964 a general radio, what is it, it's some kind of it's called the general radio experimenter, which is it's kind of like a catalogue of general radio stuff and electronic things, but they have a whole like, page dedicated to a new plug for patch cords. And it's, I don't know, it's like, showing like the wonders of banana connectors and what you can do with them. It's kind of cute, but they also call out that they had introduced the this new connector in 1924. So I don't know it's just it's funny, like when you go searching for something like I wouldn't have thought that the banana connector the humble banana connector, had such a dark past so also do you know why the banana connector is called a banana connector?

It's definitely not the scale of a banana. So it's not that

and this is what this is kind of dumb, in my opinion, but like, you know how the little the actual Jack part of it has the spring arms little the little spring. Yeah, the little springs, someone that sometimes said that that resembles a banana. So it got the name banana connector. And actually, so most modern banana connectors have one of those leaf springs, I guess you could say. Or as hard they have four of them. And the original ones had one. So it was like basically a stood with one arm on it.

Ah, that would actually look more like a banana. Yeah,

yeah, I think so. Makes sense? Yeah. There you go. All the things you never thought, or you never knew of and you probably didn't care to know.

Yep, yep. So speaking of that, I got one more quick topic. And I'll get a little weird, but we'll see this,

this whole episode is weird. The

lot, I think was last week. Anyways, Elon Musk, we don't talk about him a lot, because we don't actually talk about pretty much any one on this podcast besides us. But yeah, Elon Musk. He has a new company I never heard about before, before this press release, but it's the neural link implant company. And this is just kind of a weird sidenote, go and watch the press release. about it. Basically, they had a whole bunch of pigs and they put like a circuit board in the pigs brain. And it's not to control the pig. It's like a Fitbit for your brain is how he pitched it. And it kind of makes sense, spacey is just using a bunch of wires to read neurons, right? Kind of a cool idea. It's been done and experimented with before. But I think they're going with that looking at like, in a different way. But one of the examples that he had for the use case is they were able to predict where the joints of the pig was at. So the pig was like walking on a treadmill. And they had one of these things hooked up to his brain. And they were able to predict where the joints of the pig were at by just reading the neurons. Which was very interesting, because then he didn't say anything else about it. And I'm like, That is that is totally, totally where that application is, is making prosthetics.

Oh, like, you can know where the prosthetic is in space.

Yeah. And then your, your, your neurons would fire and it would your that prosthetic would move. And then the brain is really good at error correction. And so you know, they were like, oh, yeah, it's only like, 95%. Correct. And I'm like, that's probably about as good as what the human brain is normally good at, like body positioning.

Unless you're like an I bet you it's a hell of a lot better than 95%. Could be I don't I don't think 5% of your steps. You fall?

No, no, but your body is very good at feedback.

Oh, for sure. Yeah. Yeah.

Like, close your eyes and try to remove all your sensory inputs and then try to walk around your house.

Do it do it long enough. You get good at it. Yeah,

yeah. But most people are not.

Yeah. The biggest problem I'm seeing here is that they're they're monitoring the fitness of pigs and pigs are the one animals that shouldn't be fit in.

But I think it's a very interesting idea. I just thought it was interesting on the just the fact that he brought that up and but didn't make the application jump. It was very obvious to me what that would have been, well, prosthetics.

He didn't publicly make that that

Yeah, but the thing is, a lot of the the days, they showed off a lot of the tech, but they didn't show off any of the application of like, Why have that? I don't know if that was on purpose or not. But yeah, we'll see where that goes.

It's amazing what you can do when you have billions of dollars.

Billions of dollars. Yeah. Electric cars, rocket ships.

flame throwers. flame throwers. Yeah, that's the boring company. He's gonna bore into your skull and plant Tesla computers. at Tesla. All they do is they play they play commercials in your head until you go out and buy a Tesla car and then shoot it into space. Yep.

That's been done before. You can't do that again. Right.

Okay, we got to do something new. Nuke Mars, right. Wasn't that one of the new cars? Yeah.

Yeah. Cool. So that was the backup engineering podcast. We're your hosts Parkdale and Steven

Craig. Later one take it easy.