MEP EP#124: Letting The Loaf Out

Podcast Transcript

Host 2     00:10

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

Host 3     00:15

And Steven Craig, this is episode 2124.

Host 2     00:22

So this is the second episode with Steven in Colorado. That was pretty good for an intro, I think. Was that the Riddler?

Host 3     00:30

Yeah. Yeah, it'll be fun. Maybe we should just come up with a standard on who says what, but that that's just that's too stale and too boring. We'll just talk over each other every time now. Exactly.

Host 2     00:41

That's way too boring. That's last week, I was talking about the Mac 6682. Which was that thermistor? Like ADC chip? Yep. So I started designing a breakout board for it because I went to the you know, I went to the usual suspects, like SparkFun

Host 3     01:00

And element 14. Oh, well, no,

Host 2     01:03

SparkFun and Adafruit. I got a notification popped up on my computer. And I'm like, oh, close that out.

Host 3     01:13

Just completely derailed.

Host 2     01:17

So yes, I'm designing a breakout board. So I'll probably order that soon. And I haven't even got the thermistors in yet from China. So whenever those show up, I can actually start messing with them. Yeah. So probably the next couple of weeks or you know, how typically is never

Host 3     01:36

Well, but how it typically is you'll have everything on your bench. And then never

Host 2     01:40

Yeah, they never Yeah, I'll be like, oh, yeah, I got all the person that's like the next update. And then about eight months later, I'll be like, oh, yeah, remember that thing? It's done. I also have a came out today, Tuesday. So it's actually gonna be like, the previous day for the people listening to this podcast, I had an engineer spotlight on all about circuits. So like, I'm famous on that website for like, a day. Cool. Or infamous, depends on your point of view.

Host 3     02:19

I need to check that out. I haven't even seen that. I need to see what they what they've written about you.

Host 2     02:24

Yeah, it was kinda like a q&a. And I just answered a whole bunch of questions about, like, my background and, and macro fab and that kind of stuff. So cool. And then because, you know, we always need more projects. I picked up one of these like Florida, Ford, climate control seat modules, they usually the things that go in your, like, the fancy seats that, you know, heat up and cool down the seats. They're, you know, climate controlled. And so I've always wanted one because they look really cool. And I was really, you know, want to know how they worked. And I was just, I had basically set up an alert on eBay to alert me when one of these things comes on sale, and is below a certain price. 40 bucks, and this one triggered and I'm like, okay, cool. So I bid on it, got it. And took it apart and kind of to break down of it, and try and figure out how it works. So it uses a Peltier mm, which I kind of suspected, but I didn't really know like, when the Peltier you know, is activated, one side gets super cold. Once I get super hot, how do you get rid of the heat or the cold depending on what you wanted? If you're heating something up or cooling something down, and it basically has the Peltier sandwiched between two copper heat sinks, and then it funnels a fan over both of them and then diverts the air. So one goes into one, you know, channel would go in your seat and one just gets exhausted somewhere.

Host 3     03:56

Somewhere. Yeah,

Host 2     03:58

I actually hadn't figured that part out yet. Like, where does it exhaust it? I figured, like, my dad's Tahoe, actually vents it back into the cabin, which doesn't seem really efficient.

Host 3     04:12

Yeah, not in Houston. Yeah. And so

Host 2     04:15

I'm like, I wonder if you could just vent it outside or something like that.

Host 3     04:19

You know, the first thing that came to mind is, is an image of your Jeep, which, as you know, longtime listeners know, Parker has spent a lot of time on his jeep. So you have this kind of like, pristine Jeep, but I get this idea of like, a dryer hose, like from the back of the seat to like a hole in the side of the Jeep, you know,

Host 2     04:36

Like one of those plastic ones with a little flap. Just like a little ghetto one. I was actually looking for something like that. The mount underneath the Jeep, and so like air can come out of it. And then Eric can't go back in.

Host 3     04:51

Why don't you be one of those guys that have a snorkel on your Jeep, but the snorkel is actually just the exhaust from your your climate from that thing. It's not an air intake. Yeah.

Host 2     05:04

So yeah, the Peltier was pretty easy to get fired up, I basically on Peltier units got four wires now was the most confusing thing was like, which ones power up and basically just tried all the combos until it started pulling all the power my power supply can supply, which is six amps, it actually wants like six and a half amps total at 12 volt out. And I was doing some reading, it looks like these things actually get 16 volts. So there's like a step up. That's built into the car seats. I don't know if that's true or not, that was just I was just reading like an EPA paper on like, if then lighted seats or climate controlled seats are better than running your air conditioner in terms of like fuel efficiency. And they actually are. So like, if you do like if you're doing a short commute, where it's your car isn't going to cool down all the way inside it. It's better just to run your air conditioned seats at full blast in the run your air conditioner. And you actually get a better cooling effect on your body. fingerless if you have like a 2015 Ford F 150 that these things come out of and are in Houston and want me and allow me to like poke a multimeter into the seat somewhere. Yeah, that'd be awesome. It actually runs off of because there's no part numbers like Well, there are part numbers, but like, you search for that part number and zero results from Google.

Host 3     06:34

Wow. So is your plan to just like basically hardwire through a switch right to the battery?

Host 2     06:40

No, I'm gonna I'm gonna, I'm been designing a controller for it. Okay,

Host 3     06:45

Something that has some smarts in it. Yeah. So

Host 2     06:47

You can actually control it an OLED monitor the, you know, it will modulate the airflow and modulate the temperature. Because the other two wires on the Peltier are actually a thermistor. It took a while to figure out which kind of thermistor it is. Because that like right ambient temperature, it's like 160 170k, which is a really high value thermistor. And when you soak up the Peltier in one orientation, it goes up to 240k ohms. And you flip it around, so it gets that side gets hotter, I think, yeah, hotter. It goes down to 54k.

Host 3     07:30

That's a huge range. Yeah. And

Host 2     07:34

Looking at, basically started looking at like what normal temperature curves are and looks like a standard 200k NTC thermistor. So I'll probably just be like, Yeah, it's like, I'll have like three temperature settings on each side. So like three hots and three colds and be like, full blast, tiny bit, and then like a middle range on a thermos. So it doesn't really matter what the temperature is. It's just like, at least I can read it in like, modulate the power.

Host 3     08:04

So is the thermistor built into the unit, or is it built into the actual Peltier? It's,

Host 2     08:10

I think it's built. I think it's sandwiched on top of the Peltier. You can get to it because the copper heat sinks are like a poxy on.

Host 3     08:20

Ah, gotcha. Yeah, yeah. You would, yeah, you would want to be fairly efficient. Yeah. So I've actually never played with a Peltier before, but I've certainly looked at a couple of projects that could use them. You know, there's some guys who make some some temperature controlled refrigerators that use filters, and things. So the one thing that I don't know is like, how fast does it heat up or cooled out?

Host 2     08:46

Instantaneously? Really, it's like, bam.

Host 3     08:51

Wow, that's cool. Yeah,

Host 2     08:53

The heat is when you have this thing running at full power. It seriously one side has like liquid lava coming out of it. And the other side is like Arctic breeze.

Host 3     09:05

Huh? That's so interesting. Yeah. And isn't it if you reverse the polarity it'll

Host 2     09:10

Flip some Yeah. Yeah, that's cool. And so the control the Peltier because it pulls like six and a half amps, just for the Peltier loan. I'll get a ginormous like H bridge, semiconductor part. Like the Toshiba TV six, seven h 303. H G, which is a 10 amp H bridge driver for like eight bucks. So I'm like, okay, that will solve that entire problem. Don't have to worry about Yeah,

Host 3     09:42

Yeah. That's that makes it easy. Yeah. And then read in

Host 2     09:45

The thermistor. And then, bam, that's done. So then I started looking at the motor, because it's got a three it's a three wire motor. And I'm like, Okay, it's probably like a normal PC fan, where like, you spin it up, and then It either it pulses out it's RPM. So like you count the pulses and then you know how fast are the fans running? It is not that at all. It's actually what's really weird because I was just like I put power on the red black like you would think about it. And nothing happened the fan that in turn on pulling the white wire to power, spun the fan up all the way. But then run but then when I removed the red wire, it still kept running. What Yeah, exactly. It's really weird. So if you take a look at that schematic, I'll post it

Host 3     10:38

All yet partly Parker sent me an image.

Host 2     10:41

Yeah, my napkin drawing of this schematic. So I actually had to take the fan apart, which is actually a pain in the butt. It's got like a well, now it's mutilated, but it has a blind see clip? What? Yes. So the C, it's a C clip that goes on the inside of the bushing housing, but it doesn't have ears or anything on it. So it's like one way install. So the only way to get it out is to basically take a screwdriver, a thin screwdriver, and then pry the bronze apart, just enough to wedge it in and pop it out. So it's semi distributive.

Host 3     11:21

Yeah, this was not ever meant to be deconstructed.

Host 2     11:26

No, no, it goes back together pretty well, though. So that's good.

Host 3     11:31

So I had so what do you got here?

Host 2     11:33

Yeah, so I haven't taken apart. So it's got the, the box on the right is the driver. So the drivers in lb 11988 made by on semiconductor. It's a three phase brushless motor controller, kind of standard for these kind of motors. And I kind of just drew a box that was like Hall effects and windings because it's like, Woohoo, you know, don't really care about that. That does, that does what it needs to do the business and that's done, but it's not the end that I really cared about. Right. And so then I drew like the power input, which is why I was like trying to figure out why is it when you apply power to the white wire and black wire it powers up? Especially since when you look at it, that white wire goes directly to the gate of that MOSFET. Okay, and then it goes down to this 1k resistor that goes into pin 22.2 is VS, which is motor voltage. So that that goes directly to the drivers on inside torque pin. 21 is Vcc. So that powers up the internal, like sensing components of the chip. Yeah. And so I was like, why is it? How is that? Like, there's no way an amp is going across a one kilo ohm resistor? A 12 volt? Yeah,

Host 3     12:59

It doesn't make sense. No,

Host 2     13:01

It had to be 1000 volts. For that the work.

Host 3     13:06

That's that's usually how it works. Yeah.

Host 2     13:09

And so I took the chip off, or I took that MOSFET off. And it, of course stopped working. So I'm like, okay, so it is going through that part. And then just like, kind of like half soldered on and trying to figure out where the amperage was going through. And basically I figured it was going from directly from the gate to the source.

Host 3     13:34

What is the power? How?

Host 2     13:39

Because it's not a MOSFET What do you think? Well, okay,

Host 3     13:44

We where we started, we started with the presupposition that it was a MOSFET it's got

Host 2     13:51

A marking on it's a Deepak package with a marking, is it Oh 65886 A? And if you know what that is, let us know. Because I cannot. I've looked at like 200 data sheets, trying to figure out what that thing is, I can't find it. So it's

Host 3     14:08

A it's a SCR or a triack or a diag or something like that, you know, it's probably

Host 2     14:14

It's a bipolar BJT.

Host 3     14:19

And, well, you know, they're not going to be heating up the bass that much they're going to be dumping all the power in the bass.

Host 2     14:24

Well, only because of you wired up prom, you do. destroy me? Well, because I measured a point seven volt drop across what would be the

Host 3     14:38

Base to emitter? base and emitter, right?

Host 2     14:42

Yeah, you're right. You're right. You're right. And then yeah, yeah, I'm like, okay, that's kind of weird. That's not supposed to be there with if it's a MOSFET right. And then I started looking at Deepak automotive transistors, and you can have a bass current up to two amps on some of them. Oh, really? Oh, that's what's happening is if you don't hook up that red wire, all the power from the fan goes through the base that pjd

Host 3     15:13

Yeah, that's actually so what's funny is what's happening there is usually like one of the first mistakes you make with BJTs is you just like run gobs of juice through the base, because you plugged it in backwards or something like that, which that's actually like, there's always current flowing into BJTs base. Like that's how they work their current control device. So you want that to happen, but it's usually a lot a lot smaller than two amps.

Host 2     15:42

Yeah. So it's pulling. If you don't plug in the red wire, it pulls 1.6 amps into that base. Now, the moment you hook up that red wire, that bass current drops to like a milliamp.

Host 3     16:00

Yeah, yeah, cuz they probably has a fairly decent gain in the transistor. Yeah, yeah. And so what's, what's the voltage at the white wire?

Host 2     16:11

Troubles 12 volts.

Host 3     16:12

Okay. Okay. Well, so it's so it's dropping. Point seven times. Let me see here, one. So that's, that's 1.12 watts. That's it. That's enough to heat it up.

Host 2     16:27

And the thing is, wasn't he not because of how the sinks constructed? This is actually a aluminum backed clad board. One layer.

Host 3     16:39

Okay, so it's huge heatsink?

Host 2     16:41

Yeah. So that entire thing is like two inches in diameter. So the whole thing's a heat sink. And so it's like, I was trying to figure out where this powers How was the getting to the motor controller? And, yeah, so it goes. It gets goes straight through that BJTs base.

Host 3     16:58

Oh, that's great.

Host 2     16:59

I guess they could have fixed that if they put a, you know, a base junction resistor to like, snub that. Snap up a bit?

Host 3     17:12

Uh, yeah. I mean, it wouldn't solve that. But if we would cut it down to whatever's the resistor limited to?

Host 2     17:19

Well, I'm saying solve it as terms of someone trying to reverse engineer it without having to crack open the fan of what that pin did.

Host 3     17:28

Well, but but here's the thing you ran it with. It has three wires, and you ran it with two. So that's kind of like, well, something's gonna happen. Yeah, no,

Host 2     17:36

That's the thing is like, okay, maybe we're not my first thought was, maybe the white wire is positive power. And Black was ground ones, the red wire to and so I put my scope on, I'm like, maybe that's where you get the pulse? Or some

Host 3     17:52

Yet, See? See? Pull that up?

Host 2     17:56

No, I put that on the on the scope. And I'm getting like a constant five volts. Yeah. And so it's actually like, the reverse bias of that diode that's up there. That's for reverse polarity. It like it leaks enough to where like, the scope can pick up five volts off of it. Wow, that's crappy. Yeah. It's probably the cheapest diode that can put it on there that it's good that diodes like it's labeled for a four n. So it whoever knows what that is.

Host 3     18:26

Yeah, automotive stuff actually has funky part numbers.

Host 2     18:30

Yes. Yeah, I ran that same thing doing the Jeep radio and cross referencing part numbers from Chrysler part numbers to actual part numbers. But you can Oh,

Host 3     18:41

Yeah, I remember you were you were actually, you were doing hardcore reverse engineering by looking at just random data sheets matched the chin out that you saw, yeah. Hey, that's that works. It worked. It took you a while, but it worked. Yeah, actually, dedication

Host 2     18:59

Actually did that same thing with the lb 11988 a bit, because this is actually the HR version. But when you actually search for that part number which is on the chip, it comes up with a smaller package that doesn't match the pin out. And so I had to keep searching and eventually found the right byte. datasheet so it'd be nice to see if that's actually a BJT you know, guy right there. I'm gonna bet you it is and then it makes it smells like one. It's may make sense now. But when I was thinking like it's a MOSFET, because it's like, it's a Deepak package. The how it's laid out is correct for a MOSFET. And I'm like, How is the gate on that? MOSFET point two amps. Impossible. Unless the MOSFET is blown.

Host 3     19:48

You know, it actually kind of reminds me there's a there's a YouTube guy that Parker and I both watch. I'm not gonna actually call him out this time, but he takes apart stuff occasionally. he'll he'll pull up the circuit board. And and if there's just if there's anything that looks black and has legs on it, he's like, Well, that's a boss fed, right? It doesn't matter what it is, he's always just like, well, and if you're looking here, it's like, no, no, that's not I love it.

Host 2     20:19

I did make some guesses or in good, you know, choices in terms of figuring out for the MOSFET at first, but Further testing showed it was not.

Host 3     20:33

Yeah, it will. And if you pull that off again, you're probably not going to but you could always do you know the diode check on on the collector, the base and then the base to emitter. And if you see two diodes, then you're pretty much guaranteed it's a BJT. Yeah,

Host 2     20:46

I might do that. Again. If the motor bearing can suffer another disassembly. It's been a part of a couple times that see clips getting a little tired.

Host 3     20:57

Especially cuz you're you're cranking on it with a screwdriver.

Host 2     21:02

It's a blind see clip hole, like it's actually potted to from the get go. So you had to take the potting out to get to the C clip. Wow. Yeah.

Host 3     21:12

Well, they never expected one of these would get in the hands of a guy like you.

Host 2     21:15

Exactly. Maybe someone for that design this thing? Well will chime in, let us know what that part is.

Host 3     21:25

Where can they chime in at podcast

Host 2     21:28

At macro comm. Email, that's an email address.

Host 3     21:35

We're actually jumped on the Slack channel. There's a bunch of people on the Slack channel now. Yeah, just what Mac fab that slack.com

Host 2     21:42

Something like that. They actually were helping me out reverse engineer this thing.

Host 3     21:46

So Oh, cool. Yeah, it's there's a lot of lively conversation all day long.

Host 2     21:51

Yep. So Stephen, what lively conversation have you brought to the table?

Host 3     21:56

Not nothing, nothing at all? No. So last last week, I talked about that micro tracer, where the you tracer, whatever. So I actually did some work on this this last week, in my downtime, and so not I'm not going to like post up a schematic or anything, because it's fairly boring at the moment. But one of the things that I was wanting to do was have it such that you could, you know, flip signals all over the place within the microchip shirt, because there's, there's six separate signals that the the tracer outputs, two of them are heater pulses, so they are 12 volts, but they're PWM 12 volts. So they're, they're also high current. So you have two high current high frequency content signals, you have another one, that's zero to negative 40 volt, which is your control signal for YouTube, you have two signals, separate signals that are zero to 400 volt. And they, they, they are at zero volts. And when you press go, they jump up to whatever you choose, but anywhere between zero and 400 volts. And then you have another signal, which is effectively ground. It's sort of virtual, but it is a ground. And so what I wanted was a circuit where any one of those signals, I could automatically route that to any pin on a coop socket. Yeah. And that was originally going to do it with just rotary switches where you know, it just you kind of just like, switch whatever you want. And then I started thinking about Yeah, yeah, of course, this is total scope. Great, this is gonna be great, because originally, I was just gonna be rotaries, which is all going to be like a passive thing. And then I was like, well, but there could be the chance where I could, you know, put one of the 400 volt signals into the high powered current 12 volt thing, and I'd blow both circuits or, you know, so it was like, how do I make something that I can control such that it can never like short itself out and stuff. So yeah, it's totally scope creep into now I have like a whole board with an STM 32 processor, controlling a relay matrix, and I have nine rotary encoders that are all push button, and nine, seven segment displays. And they all represent one pin on a tube socket. And if you want to change a pin, what you do is you press the number like say number three, you press number three, and then the seven segment display will flash and you can rotate it and it'll say anode cathode screen heater and then you press it again and it'll rearrange the the relay matrix set that drives that signal to that pin on that so yeah, I know I scope creep it real freaking hard.

Host 2     24:40

But like, you know, something that like you will solder together in an afternoon to like, a week's worth of layout.

Host 3     24:48

Oh, yeah. Oh, yeah. Yeah, I totally digging it. I'm loving it. I actually have most of the schematic done. It's it's the thing about it is it's tons of Step and Repeat like you design like one little sub circuit and then multiply by nine, and you just keep doing it. So I've got almost everything done. I just need to excuse me. But the, the rotary encoders. All there. And I chose actually the the STM 32. Chip, mainly because I've never messed with STM before. So I want to learn that. I mean, I shouldn't say that. When I was working at McAfee HEPA did a lot of STM work, but that was for customers ton of customers that use them. So I'm really, really well aware on like, the hardware aspect on how to program them and how an options like that. Yeah, yeah, but but I've never actually like, you know, coded for them. So it seems it seems pretty nice. And let me see here, I actually have the port number. We pull it up. I'm using an STM 32 F 373 C eight t, which is a 41. Ah, you know, to be honest, I have no clue. I wouldn't, I wouldn't know what is or isn't popular with these things. But But this guy has a 48 pin, qf P, and it has just enough pins, such that I can shelve all the rotary encoders. And all of their push buttons, which is directly to the Yeah, direct IO. So it makes it easy. And I'm just because I am actually throwing USB on it. Because with these chips, you can put them in DFU mode and just program over USB, which makes it super easy. And I don't have an SD link. available here. Good. I'm sort of building this project with the fact that like, I'm in a temporary apartment, I have my soldering iron and my computer. What can I do with that? You know, like, how do I make that work? So like if I could program over USB? That makes it easy? Yeah. So So yeah, I almost have that done. My wife is actually going out of town this weekend. And for for a lot of people. They're like woohoo, like I get to go party. And for me, I'm like, Yes, I get to do PCB layout all weekend long. So yeah, I'll be doing that. Hopefully by next project, I'll actually have something that maybe it's even on order. Who knows? That's cool. Yeah. So that it's is scope creep. And it's a brute force, you know, solution to a problem. But given all those signals that I was talking about, you know, there's some that are high current and high switching transients. There's some that are zero to 400 volts. There's others that are zero to negative, you know, they're, they're like, you have the whole gamut of everything. How do you like, switch those to their respective pins wherever they need to go without crosstalk and being able to handle all of that stuff. And the only thing that came to mind is like, basically have to do with relays. Like that's, that's your only option. i It would be awesome to use, like analog switches or something like that. But, and at the same time, you want that you want things to be fairly precise. Like if you send it 300 volts, you want you want an actual 300 volts to get you there. Yeah, yeah. So

Host 2     28:02

I'm thinking it's like, you went from like, 60s era design where like, it will have like, a just has warning on the box. Yeah. Modern day like, hold your hand safety device?

Host 3     28:18

Yeah, yeah, absolutely. Well, and and, and this is, this is just totally a goofy project. But I'm gonna, I'm gonna send some pictures to the, to the guy who designed the actual tube tester. You know, this is like a secondary little thing. So, you know, maybe, maybe he'll like it. Maybe he won't. But I want to actually put it in a box. So I have to repackage this, too. Now, because because I, for years, I used a little project box. And it was great when I was running Craig amplification, because I use this to test tubes all the time. And that box is kind of seen better days. And there's a couple other things I want to do with it. Because I always ran this off of a power supply. And now I want to actually have a power supply inside the box and stuff. So I'm sort of doing upgrades all around,

Host 2     29:06

You just put an A C 13 power supply or power cord on it.

Host 3     29:11

Actually, so he originally designed this. I know I don't know. But the original design actually runs off of a laptop power supply. So I'm just going to grab a regular 19.5 volt laptop power supply. And it only it only consumes like 200 milliamps tops. So I'm just gonna basically buy one of the power supplies and install it inside the box. That's cool. Yeah. Yeah, it's cheap and easy. Yeah.

Host 2     29:38

C 13. That's like the PC design. That's power cord thing. There's probably a NEMA number as well for it, but

Host 3     29:48

Oh, okay. No, yeah, I see. I see what you're saying. Yeah, yeah, I didn't know that. Those were called C 13. Yeah, I always just call them IEC cables. But that's probably wrong. Well, I

Host 2     30:00

Think they have an IEC number. So it's not 100%. Correct. But it's also not 100%. Wrong.

Host 3     30:07

Yeah, I mean, I'm not going to be I'm not going to be UL listing this singular project here. So

Host 2     30:15

Spray paint UL on the side,

Host 3     30:19

Gig, get a stencil, but make it big so that people can see it. They're like, Oh, this is safe, safe.

Host 2     30:26

Bleed over into paint drip on it.

Host 3     30:28

Yeah, we have a Tormach mill at my new job. So I may ask permission to mill out a faceplate such that like all that seven segments could fit through it and stuff like that.

Host 2     30:42

Oh, you should just make it just a solid block and just make a billet case.

Host 3     30:48

It's like one solid yet, but, but like big make it like two feet by one feet by like three inches thick? Yeah, yeah. Yeah. It's like it's like a $400. Block of of, you know, raw steel or something.

Host 2     31:02

Aluminum? Yeah. Aluminum. Yeah. He's like, Hey, boss, I need to borrow the mill. For how long? Two weeks?

Host 3     31:09

Yeah, right. I'm also gonna need 50 of your endmills

Host 3     31:19

All right, cool. You want to run off to the RF? Oh,

Host 2     31:22

Yeah. And you're, you're in command of the RFO this week?

Host 3     31:26

Yeah, I got I've got two things on the RFO. So the first one is titled, digital circuits now with more analog. And then I got to like a fun little thing for you after that. So this first one, we've talked about these before, but have you? Do you remember memristors? Or have you heard of them recently?

Host 2     31:49

Ah, not until I wrote that down on this on notes today.

Host 3     31:53

Cool. Cool. So this was actually kind of a, you know, it's sort of hit like big news. Few years back HP designed the first or actually physically built the first memristor which a memristor is they call it the fourth fundamental component, you have resistor capacitor inductor, and then there's a fourth memristor, which it doesn't feel very fundamental if you don't just like if they're not just lying around, you know. But, but regardless, if there's a there's a cute little chart online that you can look at where it shows like, you know, a resistor relates voltage and current, a capacitor is charged and blah, blah, blah, blah, blah, well, a memristor. I think a memristor relates charge with a magnetic flux. I'm probably wrong in that, but regardless, it relates to fundamental characteristics together, and basically what a memristor is, it's a resistor that changes its resistance based off of the current flowing through it.

Host 2     32:55

Memristor is an electrical component that limits or regulates the flow of electrical current in a circuit, and remembers the amount of charge that has previously flowed through it. So it is,

Host 3     33:06

Thank you Wikipedia. Yep. So yeah, so I actually read a fun kind of example, where it's like, think of like a water tube, or like a sewer tube, that would grow or shrink based off of how much water flow through it. And like, if you so if you pass a certain current through, it would expand to be whatever it needs to. And as soon as you stop that current, it just stays that way. So in other words, you sort of have a two to lead ID component, just like a resistor, capacitor inductor that has memory in it. So you can actually store things in it. And because it's a physical aspect is you're not actually like storing charge or anything. It remains that way. So it's like, it's memory that locks itself in place, like even flash memory is holding on to a charge, you know, you know, I think flash memory has 1000 years what they say it'll hold memory for that long. Well, the memristors supposedly supposed to hold it for longer than that. Which is, which is pretty cool. But what the reason why I brought even put this on an RFO is I saw a an article that actually referenced an article that came out earlier this year in January, where a university has actually now developed a memristor that can record seven bits worth of data in a single memristor. So, in other words, they have a, like, think of that water tube that can expand to 128 different, you know, resistance values, and they can read that, and they can read that. So there's, we now have two legged components that store up to seven bits of data. That's pretty cool. Yeah, it's pretty so think of things like if you you know, if you look inside of like a memory You know, unit, you're, you're gonna have a transistor plus whatever passives go along with that for every single bit of worth of data. Now, the you know, with with memristors, the hope is in the future that you can store way more data in a single item. Yeah, single unit. And it'll just stay that way. You know, it's not, it doesn't have to hold charge or anything. And one of the things that they actually were discussing in this article that's on electronics weekly comm if you want to go check it out, they're discussing that they actually put two of these memristors in an output of a CMOS device, a totem pole CMOS device, which is usually just a, you know, a p channel and an N channel MOSFET that basically allow you to just switch an output higher low. But with putting two of these, you know, memristors in place, they could actually have the output of a CMOS device be an analog value in between the two channels, which is really cool, if you think about it, so so maybe he you know, in in, in our lifetime, we might be able to see chips that have memristors on both the input and output, such that you can just digitally pass an analog value, you know, without having to like, sample it, you're just literally just passing it a value. Yeah. And and so, you know, the the days of binary may be going away, you know, with these things. So, apparently, I've heard that they're notoriously difficult to actually make, and they're not cheap. So who knows, maybe somebody will crack the code, and we get something with like, infinite values or whatnot. But I thought that was pretty cool.

Host 2     36:39

Infinite values. brings us back to the potential geometers.

Host 3     36:44

Yeah, potentially ominous? Yeah, the data sheet that that has a it's a for its it, it was positional accuracy, or it was like rotational accuracy. And you said theoretically infinite, like, yeah, that's yeah, that's classic. Well,

Host 2     36:58

Remember, I'm transistors used to be, you know, super expensive and difficult to make. Oh, yeah. Yeah. So, you know, I think if there's going to be an application for these things, they will be made.

Host 3     37:14

Oh, yeah. And, and going down. And, you know, I remember reading the first article, when HP said that they, they were happy that they made four of them, that actually displayed, you know, characteristics of memory distance, you know, and now they're now they have some more, they're getting up to the point where it's like seven bits per memristors, you know, that's usable. So, and one of the biggest applications for this is Siewert. Memory, of course, yes. Yes. Memory, because this can this can make memory smaller and faster. And if you could just keep adding Bit Depth to each memristor. Just, you know, and and technically, this is a passive device, you know, it's not active.

Host 2     38:01

How does it how's the tunnel shrink? In that example? They pass it a reverse current?

Host 3     38:10

Yes, yeah. I think yeah, well, current One Direction expands, and currently other shrinks. And there's a lot of there's a lot of applications actually, I watched a YouTube video on this a while back, it's, it's confusing, but interesting. At the same time, these work really well in making memristors gates. So So gate level things, gate level components that are passive gates, that, remember their states. So you know, you pass a current into a game. And then even if you turn your device off, it's still there, you know, they also read something, you know, guys thinking into the future, it's like, your computer can just shut itself off at any time and just remember where it's at, effectively. And as soon as you want to, like fire backup, it just reads all the states and just goes to town. I shouldn't even say it's not really even reading states. You see, what's interesting about these kinds of devices is they really do blur the line really hard between digital and analog, because you have to worry about where the analog signal is. Now, in between your two states. There's no high and low anymore with these things. So it's costing? Yeah, I thought that was food for thought. Yeah. Hopefully

Host 2     39:21

In two more years, we get another update on these devices. I think we talked about him when he came out with them. I remember them once on the podcast.

Host 3     39:33

Yeah, we've talked about them before. But you know, I think it's gonna be a long time before we before we see anything that you can actually do with them. I even did a search earlier today. I was like, Can you buy a memristor like get does anyone have them? And I, you know, just a quick Google doesn't give you anything, so even to learn how to play with them, but there are there's actually already books available. that, you know talk about design work, so they must be available somewhere but

Host 2     40:06

I actually just googled memristor for sale. Yeah, and b s AFW. Eight discrete members, there's 16 Dip. What is this? Like available? Straight from the lab of memristor pioneer, Dr. Chris Campbell at Boise State University?

Host 3     40:28

Well, okay, yeah.

Host 2     40:30

Let me send you the link. Yeah, I

Host 3     40:32

Think um, is this 42? To $220? Yeah. So what you're looking at? Yeah, okay. So yeah, you can you can get, you can get some like handmade lab devices. But in other words, like, you can't go to one of the big guys and get memristors yet, you know?

Host 2     40:52

Oh, and it's not like, so if the first option is wafer 16 to one and says chip from wafer number 16. And one device one at a has failed quality control.

Host 3     41:06

Yeah, it's gonna be a while before you just get one that has a good data sheet that you can, you know, play with.

Host 2     41:12

Oh, so yeah, wait for 16 to two, all eight devices have passed quality control 180 bucks.

Host 3     41:19

You're gonna do it. You get Wait, you need a memristor in your butt. Air can air control. You need like, ultimate high tech to cool your badass. Well,

Host 2     41:30

So we have two names for that project. I like ABCC which is automatic automotive but climate control and even has

Host 3     41:41

Well I think I call it the canyon cooler.

Host 2     41:43

Yeah, I can cooler.

Host 3     41:47

Yeah. So you guys tell us which one you like better.

Host 2     41:53

If someone out there sends Steven or I a check for 180 bucks, we'll buy with this memristor and try to build something with it.

Host 3     42:01

We will buy $180 worth of beer and then later on buy a memristor. All right. So yeah,

Host 2     42:13

You can buy a memristor?

Host 3     42:16

Well, you could buy lab memristors. Yeah, you can get

Host 2     42:19

Discount if you order more, they got price breaks.

Host 3     42:24

If you order over one, yeah,

Host 2     42:27

Get a price. If you order 32 of them, you get a 30% discount?

Host 3     42:31

Oh, it's interesting, because it's probably not probably it's gonna be, you know, guys from a lab selling and the only guys who are probably buying are guys who in a lab,

Host 2     42:42

You know, exactly. And now to goofy engineers. Want one?

Host 3     42:47

Yeah. Yeah. For labs, you know, I was thinking about I was thinking about it earlier, there's got to be some really cool automatic gain control functions. By being able to morph current, or with these, I would think that there's a, that it would be possible to do conversions, like, linear to exponential with this kind of thing, where if you linearly ramp up current into this thing, because it's changing its resistance, can you can you turn that into something else? You know, like, if you put a constant current into it, what does the voltage to across this thing is? So there's a lot of, there's a lot of fun things that, you know, I'd have to look at, like, its response curves, and what resistance it's gonna, you know, flow between and things like that. Also, like, there's, there's, there's for sure, a way that you could damage these things, but I'm not entirely sure how that is, you know, and the thing that's interesting, if it's changing its resistance all the time, then it would also be changing how much power it dissipates. So I don't know exactly how that works.

Host 2     43:56

Yeah, with a finger something out.

Host 3     43:58

Yeah. It's pretty cool. Yep. Okay, so on to the next Arvo. Sure. So this one is actually a question that I had for you. And actually, funnily enough, it's sort of in a way goes a little bit with, you know, what you're saying earlier with your, the engineering spotlight that you had with all about circuits? Well, they were asking you about you and young, like, what makes you tick and things like that? So hopefully, my heart Yeah. Unless you haven't told me, but you're the only engineer at macro fab right now. Right? Yep. There's not another engineer there. So if there were to be another engineer at macro fab, there's a pretty high likelihood that you would be part of the the interview process for said engineer, right? Yeah. So what if thinking about Parker and thinking about macro factors? What would you want? Or what do you want in an engineer? For macrophylla?

Host 2     45:05

Um, I would say if I, if so if I was hiring someone from apkfab, for an engineering position it would be, can they look at a design or have they design something before from A to B from like conception to product, it's not really even like an electrical engineer at that point, you turn more into a product engineer, I would say that's what you are to Steven, you're not an electrical engineer, you're a product engineer, you can do the majority of the mechanical design work, you can do the electrical work, you can come up with the concept and price it out and make sure it's reasonably priced when you sell it. Stuff like that, which is everything a product engineer needs to do. And that yeah, that's what I look for in engineers. is cool. So it's not I don't, I don't like to hire people that have very narrow skill sets, mainly because we're a startup and like you, you have to have lots of skill sets to, you know, thrive at macro fab.

Host 3     46:10

Yeah, yeah. So so. So I think that's actually a really great. That's, that's a great answer to that. And, and one of the things that I've been told before, and I've done it actually before, is if you have a product or a device that you've made, bring it to an interview.

Host 2     46:30

Oh, yes. I really like seeing what other people are building and designing. And that's like extra brownie points.

Host 3     46:38

Yeah, especially because it makes things really easy to talk about, because it's just right there. Questions. People, the person who's interviewing you, if you're the engineer, they, they will just have questions about the device sitting there as opposed to them having to, like, generate questions, or, you know, even worse, if they get, like, a sheet of question from HR, you know, those are the worst. But you know, if you have a device actually sitting there, that's, that's, that's really, really good. So, well, there you go. If if in the future, there's an engineering position available at macro fab. Parker gave you some, some sweet inside. Inside knowledge. Yeah.

Host 2     47:21

I would say that's, for me, at least, for me, that's, I would want to work at places that are like that. Yeah, for sure. And it sounds like WMD is like that to

Host 3     47:35

Where I, I brought my synth along and brought it to the the interview process. Yeah, I brought in. It's funny, because it's a musical instrument. We didn't play it at all. In the in the interview, we just talked about it. But like, that was that was enough.

Host 2     47:52

You could have brought an empty cardboard box that says, Steve, since on the side.

Host 3     47:58

They'd be like, Please, can you explain your design theory behind this? I was like, Well, I got a cardboard box. And I got a Sharpie. And I just wrote on the outside, no,

Host 2     48:07

Two cables in it, and apply 40 kilovolts. And then get Ross. And that's the sound. And then yeah, no, and you use your mouth and you just go boom.

Host 3     48:22

So actually, now Now this, okay, because throughout my time at macro fab, I certainly got asked this multiple times. And I know you and I have together been asked this multiple times. So what we just described, right, there was a guy who probably has some experience already, you know, he's already designed a product or whatnot. How about if you were interviewing a kid fresh out of college? What, what about what would you say to a kid like that?

Host 2     48:49

What was your senior design project? And what you do want it? Can you bring it here if you can't bring it here, because it's something like my senior design project was really lame. And it was like, it was software and I'm a hardware engineer, I wrote software for my senior design project.

Host 3     49:09

That is kind of lame. Yeah.

Host 2     49:12

But yeah. I would say, like, what hobbies are they doing? I mean, you don't have to have, you don't have to build electronics to be a really good electrical engineer. But the kind of stuff I'm interested in engineers is in engineering is part of their life. And one of their hobbies as well.

Host 3     49:38

To see if they eat and breathe it.

Host 2     49:39

Yeah, I'd like people that are like that. I know. That rubs some people the wrong way, like work should be work and you go home, that's fine. When you go home, I don't work on that crap stuff. I go home. I work on my own electronics and my jeep and all that stuff, like, separate. And I understand that it's just like That's what that's those are the kinds of engineers I like to hang out with. And I think if you can't hang out with your co workers, you probably didn't get the right person.

Host 3     50:14

Parker wants to be a party all day long. Oh, yeah. All the time. Oh, yeah. Well, okay, so there's there's one other thing and this is this is hot tips by Steve right here. If you're if you're interviewing with Parker, if if if he asks you what music you like, just say meatloaf. And you get a job immediately. Especially if you can listen to meatloaf all day long. Yeah, I think I think we got in trouble multiple times because we had meatloaf going for multiple hours talkspace Yeah, it's big McBoatface that's right yeah. Now this is the macro fab reminiscing nostalgia our right here and it's only been like a month Yeah, secret hot tips by Steve for getting a job

Host 2     51:06

Or laser Hawk

Host 3     51:08

Laser hook is good yeah, the guys the guys at my work no laser hawk and countries and all those guys

Host 2     51:13

They don't know meatloaf though.

Host 3     51:16

I'm sure they do know meatloaf but they're you know I haven't I haven't let that side come out yet. I haven't been there long enough yeah, it's kind of like you get to wait a little before you get weird

Host 2     51:39

You hate the marketing people will be angry if I made that the title of the podcast for this week

Host 2     51:49

Okay, let me wrap up this episode.

Host 3     51:52

So that was the macro fab engineering podcast we were your host Stephen Gregg and probably doing a good easy later everyone

Host 2     51:59

Thank you. Yes, you are listener for downloading our show if you have a cool idea project or topic or meatloaf song you want Steven Knight to discuss. Tweet us at Mac crab or email us at podcast at Mac fab.com Also check out our Slack channel. Everyone there loves meatloaf, especially the food. If you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest episode right when it leases and please review us wherever you listen. So it helps the show stay visible and helps new listeners find us

Transcribed by https://otter.ai