Podcast Notes
- Parker designed the MacroDuino! Basically an Arduino clone using Type C USB, a FT230X as the USB UART bridge, and proper power switching with the Ti TPS2113A. (See Figure 1). Also supports 3.3V and 5.0V.
- Earlier this week Parker had an article about adding external Part Libraries and DRC rules to Eagle. Check it out here.
- Some new work on the PinHeck Pinball System. Parker is working on adding the Octavo OSD3358 to replace the Parallax Propeller as the A/V processor.
- Stephen has been working on his Brewing Rig. He is giving a talk at Destination Codes about the Brewing Rig since it will involve programming and hardware.
- For flow control, Stephen is using a Atlas Scientific flow meter and a motorized ball valve from Amazon.
- LattePanda is the platform Stephen wants to use for the Brewing Rig. Runs Windows10.
- KAIST process engineers at the Korea Advanced Institute of Science and Technology have developed a chip packaging technology that creates flexible integrated circuits. The substrate they designed is very interesting.
- Fixing the Ampere: Redefining the SI Unit.
- ON cleared to buy Fairchild. ON has received the US FTC’s blessing to proceed with its $2.4 billion takeover of Fairchild on condition that it sells off the Ignition IGBT business.
Special thanks to whixr over at Tymkrs for the intro and outro theme!
About The Hosts
Parker Dillmann is MacroFab's Co-Founder, and Lead ECE with backgrounds in Embedded System Design, and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. He also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components. Parker graduated with a BS in Electrical and Computer Engineering from the University of Texas.
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.
Podcast Transcript
Host 1 00:09
Hello, and welcome to the macro fat engineering podcast where your host, Parker, Dolman and Steven Greg. So this week been working a lot more on internal stuff, right? Yep. I got the macro Duino. This kind of like a side project at home I've been doing. And it's basically as the name implies, it's a Arduino with macro fab all over it. Right? And the state of Texas and the state of Texas on the back of it. It's pretty cool silkscreen design on it.
Host 3 00:39
You put a little bit of your touch into the circuit, right?
Host 1 00:42
Yeah, so it's not a direct copy. It uses USB type C, Mm hmm. It actually uses an FTDI USB bridge.
Host 3 00:50
And the power switching,
Host 1 00:51
It has a really nice ti power. mux. Yep. TPS something something something I don't remember what the number was right. And that's actually really cool because it just senses which powerline is being used. So it prefers to use, you know, power in verse USB. But if it doesn't have power in it switches over the USB and actually can do it hot swappable. So if, if your say your battery pack dies, you can still be powered over USB without losing your device.
Host 3 01:20
So it's an upgrade? Yeah, it's
Host 1 01:22
An upgrade. And I think it was $40 for singles. One of them.
Host 3 01:29
Hey, that's only $5 More than an Arduino. Yeah.
Host 1 01:32
And if you if you make 1000 100 of them, it's like 25 bucks. So pretty good. thing. Oh, yeah. And it uses the new 80 Mega 320 P. B. And so just a regular P. Right. Yeah.
Host 3 01:47
Cool. So that you have all those files available? Right?
Host 1 01:51
Yes. Or they will be they will be? No, they're actually already in the macro fab macro articles GitHub account or GitHub repository. Okay. And the only difference is the P B doesn't have any of the stuff in it yet because it doesn't work with stock Arduino. I still have to work on the boards, like configuration files stuff. The network by need the board. The work on that stuff.
Host 3 02:18
Yeah, right. Cool.
Host 1 02:21
I had an article come out today or not today. Was it two days ago, Tuesday about adding Eagle crap like DRC. And that's actually
Host 3 02:34
Here on our on our notes here. Article about adding Eagle crap.
Host 1 02:38
Yeah. They say how to add like external libraries, excellent. DRC rules, all that good stuff. There's an article about it. It's, you know, it's pretty simple stuff. But some people don't know how to do that.
Host 3 02:50
It's really dry too. Yeah. You want to go learn go here? Yep.
Host 1 02:54
And I've been working on a new pin heck, board, rev eight. So we're gonna drop the parallax propeller off of it. When parallax propeller does all the AV stuff audio video, we're actually going to be moving the audio video over to a Linux system. So we're looking at using the octavo Oh, SD 3358. Right. Yeah.
Host 3 03:18
Doing doing a system on chip. Yeah,
Host 1 03:20
It'd be a system in package. So si P Yes. So we'll be dropping this big BGA 40 package on there. And then we'll be able to do all the audio and video over HDMI. At 10 ADP 60 FPS you know, awesomeness. Couple Gosh,
Host 3 03:38
I don't know how many all using the pic 32. But you just dropped in the Prop will not swapping the prop dropping the prop out for this. The pin heck is a good definition of feature creep. It just keeps creeping well,
Host 1 03:51
Because it's on Revision eight now. Yeah. And the thing about it is when we first designed it, it was designed to run just an LED DMD display. Do lights so much. And that was in some audio. Yeah. Now it's gotten to the point where like, it's currently rev seven drives a LCD display. Very low, low resolution, and very slowly it drives at like 28 Hertz.
Host 3 04:18
Very slow. Yeah, yeah.
Host 1 04:20
And so we want to buy Yes, but it works. And so we want to up this up the game for next year. And so yeah, that's what this Linux system is going to do on the board is enable us to do 1080 p 3060 FPS screens with HDMI. And basically, that solves the display problems. Sure. And then it will also do audio, the audio part. So we will do like 44 kilohertz probably like bazillions numbers of you know, channels of audio. Yeah, yeah, yeah. But yeah, we're also using the pic 32 Still because the problem with the Using a Linux system is it's not real time they make real time Linux and all this other, you know, mumbo jumbo but doesn't really work too well.
Host 3 05:10
It's still good to have a microcontroller. Yeah, that that communicates with the peripherals. Yeah.
Host 1 05:14
So they actually what it is your game code runs on the pic. 32. Right. And the pig through to actually tells the Linux system, what video is the play you want
Host 3 05:24
Audio to do? Right? Yes. Right. Yeah, that's, that's a lot easier to have that direct connection to a solenoid or a switch or exactly, flipper, or whatever.
Host 1 05:34
And we're actually going to make it so that there's gonna be a USB port on it. So you can plug in a keyboard and mouse. And so you can develop your game on the pinball machine on the Linux side, and then push the code over to the pic. 32. That's cool. Yeah, so it'd be so you can edit
Host 3 05:51
That way. It would be awesome. If you could program it from the side flippers and the ball shooter.
Host 1 05:58
Yeah, that'd be insane. Like this character scroller. Exactly. Start button is like the Enter. Yeah, man. That would be brutal. Be like a good easter egg?
Host 3 06:11
I could, but I could see somebody actually doing that. Just because they could, you know,
Host 1 06:16
Well, it's like the old way of programming computers, where you would, you had a bank of eight switches, right and a line advance switch. And you would flip your your byte in line advance and keep going and you punch in on, you know, 200 or so lines of code, and then hit run, and it would calculate and run your code. Right?
Host 3 06:38
Right. But but most of the time, you would write out your code by hand beforehand on paper, and you would debug it yourself by hand before you enter it in. Because you could be at, you know, line 160 something and flip the wrong switch. And it's like shit. Oh, cool.
Host 1 06:58
So that's what I've been doing this week. Yeah, so
Host 3 07:00
I've been doing a little bit of, or a good bit of internal stuff to working on a bunch of process documents. So flowcharts, and things like that kind of, not a lot of engineers. Yeah. Yeah. They, they don't necessarily tell you, you're gonna you're gonna do this in engineering school, but flowcharts and all that good stuff and process definition. And it's all part of manufacturing, but not necessarily the fun stuff. On the other side, I've been working on my Bruin controller, which that is the fun stuff that people might want to hear. Yeah. So I'm actually going to be given a talk at a conference coming up this December. Destination codes.
Host 1 07:39
So what is destination code?
Host 3 07:41
It's so actually, in Episode 20, we had Matt keys and Brian Dorton, from the iron yard on Matt keys is actually he set up this conference, which he actually calls it an unconference. So it's basically a gathering of like minded individuals who are interested in coding and electronics and just a whole various slew of nerdy topics. And I'm actually going to be giving a talk there about hardware, electronics, and how they relate to a cool topic, such as beer brewing,
Host 1 08:12
Sounds like a conference, it's sort of,
Host 3 08:15
It's a conference that involves a couple of days of skiing, and you know, good food and interesting people and all that stuff. So it's less of like, go to this room in a hotel and
Host 1 08:28
Business. But instead of, you know, business dude going golfing, yes.
Host 3 08:34
But so so in preparation for this, I'm actually developing an electronic brewing control system that I'm going to be presenting some portions of. But so so brewing is inherently a chemical process. I've been breaking it down into individual items that you can measure. So when it comes down to you know, brewing, it's, it's working with grains, it's working with water, it's working with a couple of physical processes like temperature, you have to maintain temperature, time, flow rate, a handful of things like this. So so I've been the biggest one actually, is flow rate for me, that I've been kind of toying with at the moment. And it's actually really difficult to find flow sensors, or good flow sensors, I should say,
Host 1 09:25
You know, inexpensive flow sensors that are good.
Host 3 09:29
They there we go there, right, because I don't want to spend $1,500 on a flow sensor. Yes. And there's nothing like beneath the industrial world. There's not really there's no sir. Yeah, don't get me wrong, they exist, but most of the time, they're like, do you want to measure, you know, 50 gallons an hour? Not really, I want to measure half a gallon a minute, you know, I want to measure like real small flow rates. And that's been difficult to find, but I did find one on Atlas site. If ik, which they have a bunch of really cool environmental sensors, pH, color, pressure, temperature, they have a whole bunch of different sensors that they offer. And they have some flow meters that I actually picked up one. And I've been playing around with that, which has been interesting. Just because flow is so difficult to measure quantitatively through Yeah, electronics. So the, the one I got can measure down 2.1 gallons per minute. And I've been playing games with an Arduino connected to that. And I've been getting good results with it. But just measuring flows not enough. I actually purchased a motorized ball valve off of Amazon. That
Host 1 10:43
You guys got that in today. Was it yesterday? Yesterday? Yesterday? Got that in? Yes, sure. Three today.
Host 3 10:48
Yeah. Right. So we've been playing around with it, which is cool, because I think I'm going to what I want to do is I want to write a PID loop that controls the ball valve that is on the output of a pump that pushes through my flow meter. And then I'm able to basically identify and control everything. And right now, I've been doing everything on an Arduino because it takes like 10 minutes to write code to do this stuff. And if you screw it up, it's like, okay, whatever, I'll just change the variable.
Host 1 11:14
You should use a macro, which we know. You know,
Host 3 11:17
For 40 bucks. I might have to do that. Yeah. Although, okay, so all of this said, I was watching Eevee. blog just the other day, the the mailbag, which is a fun segment.
Host 1 11:28
That's the one that's pretty much the only one I can watch is what you're about to
Host 3 11:33
Say there are something like, not that bad. But well. Okay, so the mailbag is fun, because it's just a whole bunch of random crap. But he had an I didn't I hadn't seen this before he, Dave Jones was talking about the little panda, or the sorry, latte Panda, latte pan latte Panda, which is a single board computer like a Raspberry Pi, or BeagleBone kind of thing. But it runs a full copy of Windows 10. And it has an Arduino built into it.
Host 1 12:04
Now does it run? It's a can't be x86 windows 10. You know, I don't know, I don't know which shot to be the, like, the windows 10. It's running on the Raspberry Pi. It's got to be the
Host 3 12:17
Is there one that runs on Raspberry Pi? I think so. Yeah. Hmm. Okay. I don't know. Okay, so maybe it's maybe it's stripped down. From what I read. It seemed like it was a full, full copy. But but it's, it's cool, because it's just natively for me not being a programming dude, and not really wanting to dig through all the Raspberry Pi garbage. I'm like, oh, that sounds cool. So I might, I might start playing around with that. But it's just cool that that you have the microcontroller directly connected to a computer, sort of like the pin hack now is going to be with the Linux in the pic. So I'm going to start investigating that and seeing what that can do for me in my burn rate. Now the great
Host 1 13:00
Thing about the PINet because we have to basically hit a price point. It's one of those because the thing about your brewing equipment, you don't really have a price point.
Host 3 13:08
You just nailed it. No, no, I'm not looking to sell this. I'm not looking at manufacture. Yeah, but
Host 1 13:12
The Pinhead because like we have to hit X mountain dollar mounts, right. And so that, you know, we can build pinball machines and they're profitable.
Host 3 13:20
Exactly. Yeah.
Host 1 13:23
The thing is, we're axing an $8 part and put a $40 board on it. It's I got to figure out how that where can I cut out of the bomb? Get that $40 Part make it the the board the same price? It might be the point said the board has to go up a bit. But you know, it's fine, because we're also getting a ton more feature out of
Host 3 13:42
It. Yeah, as long as you get the value out of it.
Host 1 13:45
Yes. So long as we can get basically because that's the thing with the prop is we're maxing it out. It's pretty much done. Where we're at the everything's in assembly now.
Host 3 13:57
Wow, you've gone that low. Yeah, just to get into all the speed out of it
Host 1 14:01
All the speed and all the RAM we can get out of it. Oh, yeah.
Host 3 14:05
Assemblies cool for about five minutes. And then when you have to write something big, you're like, Oh, God, this is horrible.
Host 1 14:10
Our entire audio driver is in an assembly. Who wrote that? Rory elfin? Really? Yep, cool. Hey, Rory. He listens to the podcast, so, okay. Yeah. Hey, how's it going, bro? Ah, so I think that's it, right. Yeah. Yeah. So the, this is a really interesting you found this article? Yep. The is it cased case process engineers, which is the Korea Advanced Institute of tech science and technology. Yes. They have developed a new kind of chip packaging for ICS Yeah, that basically allows them to make flexible chips, flexible die yet flexible dyes. So they put on flexible circuits or flexible PCB substrate. And yeah, so you have an entirely flexible piece. b, which is really cool. Um,
Host 3 15:02
Well flexible, like silicone, basically.
Host 1 15:05
Yes. And so how they do that is basically just cut the silicone dyes really, really thin. Yeah,
Host 3 15:11
Right. Anything's flexible and thin. And
Host 1 15:14
Yeah, so that's the same thing on that and they cut to 100 nanometers. I think what? Yeah, that's what it said. And then they attach it and how they attach it to the substrate. is I got a look at my cheat sheet real quick. Oh, yeah. It's called ACF, which is the end this trophic conductive film. What? Yes. So it's all this is a substrate material that's basically glues the dye down. Okay. And it does. So it's, it's a conductive film, but it's only conductive one direction. Like a diode. No more like polarized sunglasses. Hmm. And so electrons can't go sideways in the substrate, they can only go through the substrate. And so when you glue a dye on top of it,
Host 3 16:06
It's got a weird crystalline structure. Yes. Yeah, that makes sense. Yeah.
Host 1 16:10
And so when you glue the substrate down, you the basically the this basically places the bond wires. Yeah. And so the. And so the base of your electrons flow through the ACF material and just go straight down, and they can't interfere with the next connection. That's cool. Yeah, it's really interesting. I wonder
Host 3 16:32
How much how much isolation each each has, like, is it? Can you get like 1000s of volts worth of isolation?
Host 1 16:41
That would be cool to die. And they're talking about flash memory. So it's probably only like five volts.
Host 3 16:47
Yeah, Max. Saw really cool stuff, though. Yeah. And they were talking about rolling in onto things.
Host 1 16:55
Yes. So yeah. When they it's almost like printing a newspaper. When you look at like the big drum. Yeah, it's like, spinning it putting the ink on. It almost looks kind of like that's what they're kind of explaining how this stuff gets applied to the how the dyes get put down and stuff,
Host 3 17:14
Right? Yeah. Kind of like the way a laser printer. Yeah, dust. Yep. But laser printing silicone. dyes. That's
Host 1 17:22
Cool. Yeah, it's really cool stuff. I'm assuming really exciting to see where this kind of technology goes through. Because the first when you start reading, you're like, oh, they just cut it really thin. That makes it flexible. Like no, the magic is this ACF. Conductive film. Yeah, having
Host 3 17:37
Having it such that it only flows vertically as opposed to horizontal
Host 1 17:41
Because that gets rid of your bond wires and bond wires. Why stuff can't flux do? Well, the bond wire just breaks.
Host 3 17:48
Well, there's that and the bond wire can only carry so much current. Well,
Host 1 17:51
This stuff can't carry that much current. This is for low voltage, low power stuff.
Host 3 17:55
Yeah, I get okay. Yeah. No. Cool stuff. Awesome. So I actually saw a really cool article this weekend. This was this was something I didn't know. Our girl on Hackaday. If you didn't know, yeah. Oh, my gosh. But fixing the amp in the ampere. Everything? We do don't Didn't you know that? No. So the amp though the unit of measurement for current is not fixed. It's not necessarily a known quantifiable unit. But my multimeter does know. But it doesn't work. And this was I didn't know this. And I probably should have. But it's it's it's there seven base units in the SI world. And the only unit of the base units that are derived for electricity is the ampere, but it's not fixed by anything yet. The actual definition on I pulled this directly from Wikipedia, which is the most reliable source of information on the internet, right? But okay, so get this. The, this is a direct quote, the amp is a constant current that will produce an attractive force of two to the negative seven newtons of force per meter of length between two straight parallel conductors of infinite length, and negligible circular cross section placed on one meter apart in a vacuum. You can't even do that. No, it's not possible. I can't even understand what that is.
Host 1 19:37
First of all, those two things about that you can't do. Yeah, one pure vacuum.
Host 3 19:44
Infinite length, and negligible cross section. You can't do that three things, because three things, three things. So so our entire unit of all of electricity is something that is just a
Host 1 19:57
Thought. It exists in a mathematical form. on a chalkboard, exactly. So we actually six one on one, we actually back in one, we back
Host 3 20:05
Derive our main unit of electricity, the amp, we back derive it because voltage, we can measure voltage as difference of potential in current, or a difference of
Host 1 20:19
Charge. And then we know the resistance, and then you get ampere, and then you come back,
Host 3 20:23
Do that through. So you have to use multiple equations to get to your base unit. So all of this to be said, That's ridiculous in today's day and age, because a amp is actually defined as one coulomb worth of charge passing through one unit of space, a node in one second, yeah, right. So one coulomb of charge, we actually don't even really fully know what that is in terms of elementary charge of electric electrons. But we actually have the ability to count the number of electrons that pass through something. Yeah. Because we, there's these new devices called quantum mechanical gates.
Host 1 21:08
Okay. I was hoping, no, it is an electron, just be called electron counter. Wouldn't it
Host 3 21:13
Be awesome? It will, okay. In effect, it is. So So now, instead of fixing the amp, based off of a back calculation from a volt, which in that in itself is a back calculation of a Coulomb. Now we can physically count how many electrons are going to be going through a particular area, one by one in a unit of time. Therefore, they're poor, fixing the amp and finalizing the seven, or the one electrical base unit of current, which I find this stuff super interesting that we're like, it's 2016. And we're still doing this stuff. Yeah. You know, so how far off are we? How far off of what, what was our,
Host 1 21:54
Our current estimates verse, what these things actually spit out these quantum gate?
Host 3 21:59
Oh, so So the final numbers are not supposed to. So this is all being done by NIS t, which is the International Science measurements, and they do the stuff like defining the kilogram, and crap like that. So this is all supposed to be like done in 2018. So we don't we don't have an answer yet. But one of the reasons why we this hasn't been done particularly accurately, I should say
Host 1 22:27
Humans can't see electrons moving.
Host 3 22:30
Yeah. But we also don't know the elementary charge of an electron to its infinite decimal place. So we're always wrong by a little bit. Yep, we always have some error. And so by counting the singular number of electrons, we can get to within a few parts per billion accuracy of what an actual amp is. So the in all said and done, this isn't gonna affect you know, this isn't gonna change your multimeter.
Host 1 23:01
Overnight, your multimeter is not wrong.
Host 3 23:04
No, overnight, all of electronics are wrong. No, it's not going to
Host 1 23:08
Be that Ohms law is not, you know, does not work anymore.
Host 3 23:13
Well, but so the thing is, it's the same thing with the kilogram because we're, we're NIS T is shooting to redefine the kilogram,
Host 1 23:22
It just sounds like the board at this point. Now, well, just the other day, fix amperage, or come up with an eighth SI unit.
Host 3 23:32
Just make up a new year, new year for something. But the kilogram right, was previously defined as, what is it one liter of water at four degrees
Host 1 23:41
Celsius, one milliliter of water at one milliliter? On a gram, right?
Host 3 23:47
No, you're right. I'm sorry. I'm sorry, gram. gram is one milliliter at four degrees Celsius. Yes. Which is waters most dense is dense, whatever point. And that sucks in itself, because it relies on temperature. And it relies on volume in order to find weight. But if you redefine the kilogram, like there's, it's been proposed as a certain number of silicon atoms in a sphere, it is now defined not by something you have to measure. It is defined by a physical number. Yeah. Which is exactly what's happening with the amp. We're trying to define it by a physical number of electrons passing through a space which, awesome, sweet and that's what I have to say about
Host 1 24:35
That. Awesome. Ah, not as intellectually probIem. On semiconductor, yes, is actually really interesting. Um, cuz I actually asked, I think this happened on Monday. And I asked even I'm like, so who's buying who today?
Host 3 24:58
How many podcasts have we had where We talk about someone buying somebody. Almost every other one. Yeah, it's got to be a handful.
Host 1 25:04
Yeah. So I asked you like so who are the people who are left? Who is buying who? Because it's one company that makes sense that's getting either bought or buying. But the other one on semiconductor You don't really hear about too much. Yeah. You just like happen at least in my mind. You just happen to use their part sometimes.
Host 3 25:26
You realize, Oh, dammit, I have to use an onset.
Host 1 25:30
Never thought about it. Yeah, like, because they're not like that. I guess I don't have like a presence in like, in media or anything.
Host 3 25:37
By default, you don't just go to on semi?
Host 1 25:40
Yeah. It's like, it's like, they just don't have immediate presence. Like usually no, of TI because ti market has marketing. Right? And stuff like a TI
Host 3 25:51
Does everything. Yeah. Also, how many times have you mentioned that Linear Technology? They make great stuff, but you can never buy it? Yes. You understand me doesn't have that kind of thing. Yeah. It's just like, oh, this is a non part. Okay.
Host 1 26:05
Okay. Yeah, whatever. It's really good. ldeo Sure. Yeah.
Host 3 26:09
Thanks.
Host 1 26:11
So and I didn't know how big they were at all. Yeah. The by in Fairchild. Which is like the granddaddy of all semiconductor.
Host 3 26:20
Yeah. Fairchild started it? Yeah. Way back. There's some really cool YouTube videos about Fairchild talking about some of the first ICS Yeah, and and showing their like, slides and films for ICS where they're like, we have over 60 transistors in this. It's good for you. Go watch them.
Host 1 26:41
Yeah, well, I will have to post them on. Yeah. Because I haven't seen those yet. They're cool. Um, but so on is been cleared to buy Fairchild. It's gonna be $2.4 billion. Of course, semiconductor companies have so much money. Yeah. But on the condition, this was interesting, because this is one of the first ones that has that condition that Fairchild has to sell the IGBT ignition line to I think, is basically on semiconductor. They make IGBT ignition switches. And I think fair trial was probably the only other people that do that. And so it would be a monopoly. And so fair, Charles got there. This line to make this work.
Host 3 27:32
I was about to say, if you're buying them off, why would you want them to sell their stuff? If their competitor? Yeah, you're right. They're they're avoiding much bigger problems. Yeah, in the future. So I'm
Host 1 27:43
Always wondering, like, what if no one buys it that line? Well, why don't you buy it? Sure. $1. There's my offer. Yeah. Anyways. I didn't know what an IGBT was. Something like a political term. Yeah. Does, or some kind of like, Yeah, whatever. It means that insulated gate bipolar transistor, and so I basically read the Wikipedia article on IGBT. See, we
Host 3 28:13
Both go to the world, the internet's most reliable site information.
Host 1 28:18
And basically what an IGBT is, it's, well, first, the bipolar transistor existed. Yes. And then the MOSFET came out, right? Then the IGBT. The IGBT is a combination, bipolar transistor from the low level the bipolar transistor with a MOSFET. Gates. Yeah, on the front end, right. Basically, it combines the awesomeness of how easy it is to drive a MOSFET because it's voltage driven. So the current driven, that's awesome. But you also get the insane amount of current and voltage that you can pass through a BJT JT Yeah, with that. Cool stuff.
Host 3 29:02
So you got Okay, so a BJT is a current controlled current device. A MOSFET is a voltage controlled voltage device? Correct. And an IGBT is a voltage controlled
Host 1 29:14
Current device? Yes. Awesome. Yeah. Kind of funky.
Host 3 29:18
So then, is there another one that is a current controlled voltage device? Mm hmm. I mean, you can set up a you can set up a, an op amp to do that. Yeah. But I don't put like a singular component like
Host 1 29:30
A typography. Yeah. I don't think so. Hmm. There probably is a design but there's no need for it. Yeah. But it's cool. I didn't know about these things. And basically because I don't have to play with 300 amps and 1000 volts
Host 3 29:47
And drive it from five five volts or five volts. Yeah, right. So they use them a lot in like industrial switching. Like if your motor like you need to control a monster lathe or something like that. You Use it these guys
Host 1 30:01
When the thing about ignition beat IGBT that's what the line that Fairchild has to sell right is the basically what they're doing is using the voltage control so they can limit how much current goes through the the ignition circuit on their cars. And so you don't basically have an insane in rush on some of these starters. So this is the interesting thing. I found I one of my mechanical engineer friends when I was in college, he was doing like, at like an engine class I guess or something like that. I had no idea when Macquarie schools Yeah, gear school piston class. But he was saying that with the new emission regulations for cars, yeah, you basically have your car has to start the engine has to start within the first hour of revolution. Or you already put out too much unprocessed gasoline to your your exhaust because it's not firing. It's just spitting it all out. Yeah. Because it's not burning it within one revolution within one solution.
Host 3 31:12
I guarantee you my truck does not start.
Host 1 31:15
A jeep doesn't either. Jeep starts pretty quickly, but still takes
Host 3 31:18
Things no one know.
Host 1 31:22
What's why all new cars have a start button. Because you press the button, it automatically knows when they can turn off the starter. Gotcha. Now. It's way faster than the human can turn key. Right and control. Yeah, so anyways, I'm gonna bench Thanks, Obama. Betcha these cars have monster starters. So that you get that motor up to speed like as fast as it can they just rip it. Okay. Rip it as fast as they can. Yeah, so you need an IGBT it's gonna have this insane inrush current on that motor on it make sense? Bam.
Host 3 31:59
Because Because a typical starter motor is like 200 amp or something like that. Yeah, something like that. They might be doing like 800 now yeah. 800 or something like this just completely tearing it up.
Host 1 32:10
Yep. Cool stuff. That's that's just speculation, but maybe someone will set a straight on Twitter, but I'm gonna butcher that probably. That's probably somewhere in the realm of why, huh.
Host 3 32:21
Backblaze wants to sell that because they do they already have some kind of Yeah,
Host 1 32:25
On semiconductor already has a ignition IGBT line. Gotcha. And so I'm going to betcha Fairchild and on semi are like the biggest two players. And so you have to sell it off. So you're not a monopoly anymore,
Host 3 32:40
Right? Because that's way worse than just competing against them.
Host 1 32:45
Exactly. Yeah. Interesting. Yeah. And so that's gonna do it for this week's macro engineering podcast. Yep. Yep. We were your host, Parker, Dawn and Steven Craig. Later guys, take it easy.
Transcribed by https://otter.ai