MacroFab's Misha Govshteyn and Chris Church check in with Parker and Stephen to give his take on supply chains, nearshoring and reshoring.
Why is there such a disconnect between component datasheet drawings and EDA footprint layout tools? Stephen and Parker dive into this on this podcast.
This week's topics are: Porsche's Synthetic Gasoline, Record Chip Manufacturing Sales for the year 2022, and the Raspberry_Pi Social Media Firestorm.
Parker is an Electrical Engineer with backgrounds in Embedded System Design and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. The following year he designed and produced an Atari 2600 video mod to allow the Atari to display a crisp, RF fuzz free picture on newer TVs. Over a thousand Atari video mods where produced by Parker from 2006 to 2011 and the mod is still made by other enthusiasts in the Atari community.
In 2006, Parker enrolled at The University of Texas at Austin as a Petroleum Engineer. After realizing electronics was his passion he switched majors in 2007 to Electrical and Computer Engineering. Following his previous background in making the Atari 2600 video mod, Parker decided to take more board layout classes and circuit design classes. Other areas of study include robotics, microcontroller theory and design, FPGA development with VHDL and Verilog, and image and signal processing with DSPs. In 2010, Parker won a Ti sponsored Launchpad programming and design contest that was held by the IEEE CS chapter at the University. Parker graduated with a BS in Electrical and Computer Engineering in the Spring of 2012.
In the Summer of 2012, Parker was hired on as an Electrical Engineer at Dynamic Perception to design and prototype new electronic products. Here, Parker learned about full product development cycles and honed his board layout skills. Seeing the difficulties in managing operations and FCC/CE compliance testing, Parker thought there had to be a better way for small electronic companies to get their product out in customer's hands.
Parker also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Special thanks to whixr over at Tymkrs for the intro and outro!
Welcome to the macro fab engineering podcast. We are your host, Stephen Craig and Parker Dolman. This is episode 265.
So I have an update on the resealing the 12 ounce cans project I had talked about last week, I got to test it in the environment, which is outside, right.
Yeah. And a pressured in a pressure environment, right? Yes,
yes. It failed. I actually it didn't work. It worked until you shook up the cannon a lot. And then there was enough pressure to push past the seal. And so like actually the thing like you could seal it up and just like sit there and like, you could like flip it upside down, and it wouldn't leak at all. So people who don't know that Miss last week's episode, this is like a, I was coming up with a way to reseal a 12 ounce coke can basically that when that you know, you pop the top, the metal metal seal. Who knew there's a lot of pressure inside those cans, like a lot. It's like 40 psi. Yeah. And so, but you're thinking like it's pushing against a little tiny metal things like, the lid is like, you know, that's probably half a square inch, so you only got 20 pounds of pressure. But now it's pressing up against the tire surface of the top of the can. Effectively because that's how big the seal is. So you got like, probably 100 pounds to deal with of force. And it would just push past the seal. So it would just leak everywhere. Good things that tested it before putting it like in my backpack. So I knew it wasn't going to work
well and tightening it more didn't fix it. Right.
Well, I tighten it as much as I couldn't think it go. Oh, I bottomed out. Yeah. Okay. So the next step is to good I want to see if I can make this work is I think I need to find a softer sealing surface. So that I you can put more grip more on the lip, basically, of the can. Because I'm currently using a shore 55 a neoprene rubber sheets. So I'm hoping to find something softer than that. It's actually it says it's shorter 55 A it, it feels a lot stiffer than that. Because like, it's really hard to squeeze your fingers,
you're meeting to the threads that are formed into the lid of this.
Oh yeah, I'm using like a Yeti. vacuum canister thing as like the vessel that holds the can,
will write in the end, the tooth of those threads is pretty huge. It's It's massive. It's not like a real fine pitch thread. Like it's an enormous thing. So I'm almost thinking if you were to chop off the bottom of your lid. So in other words, bring your lid up just a little bit. Then as you twist it on, there's no chance of it bottoming and bottoming out before the seal is made. Make it such that like the seal is the master of the whole thing.
Yeah, yeah. Yeah. So that's the that's actually what I tried, like I'm gonna print tonight is I'm just going to move that surface down. And just try the shore 55 neoprene again, and just like see if I can you know what actually I could do. Instead of doing that is I could just put a spacer at the bottom of the can just lift lift the can. Yeah, well, we'll just experiment instead of having to print a whole new lid. Yeah, he's experimenting by moving cardboard down in there. Yeah, is put something that's non compressible, like move it up half a millimeter. And see if that makes the seal work. Because it does feel like it does bottom out. Like you could like that seal could be compressed more.
Right? Right. You just you basically need to preload the top of the can into the material seal. Yeah.
But I'm pretty excited like it actually like did seal it until like you shook it up. That's pretty impressed. That rev one kind of worked. But I have some I have some. I don't know what images I shared on the on the for last week. But here's like, here's here's actually read one.
Oh, gosh, he's got giant teeth. Yes. It's
got like, it looks like I call your Yeah, it's like a cog, actually, it's good way to put it. It's a cog on top that you can really get your fingers on to like twist it on. Yeah, I realized that wasn't really necessary. And like it actually kind of like eats your fingers up. It's actually not pleasant to use. I'll take a picture that for the for the podcast, but then I made a like this turned out amazing. I printed neuros Look at that.
It Yeah, I'm looking at it right now. And it frankly, it looks like black anodized aluminum
over you know the only thing that when you actually screw this onto the the Yeti koozie I guess it actually looks like machined aluminum until you look at the top. And the top looks like the top looks like just, you know, it's the surface of the bed on the printer, right. And this is this is Rev two rev threes got like a, a nice Phillips on this edge here, the top edge to like, doesn't eat your fingers up. But yeah, like the neural like, works amazing. And actually, it works really well. With your gloves on too. And like you could you can get a grip, and you could crank it all the way down. Because that was never a problem. I have
an idea. With this, you could put a small like spring plunger in the bottom, such that as you tighten it down. It's always applying a constant pressure upwards,
constant pressure up. Yeah, you would need to get a taller koozie though.
Or just have the spring be really stiff and not have much of a throw on it.
I guess you can make the lid taller to to make up whatever thing you put in the bottom. Let's just try raising it up like,
yeah, just put something hard down in there. And yeah,
I'm actually gonna see if I have any plastic, like half millimeter plastic and just cut a disc out of it, and slam that in the bottom and see if that actually helps it. Or, you know, I should I should just put some of that that rubber ceiling stuff at the bottom to I'll figure something out. We'll see if that works. Because that's actually the easiest way to see if it works or not is let's keep increasing that preload until like, until we realize it just won't
work. Well. Okay. And I saw the threads inside of the lid. I think we talked about this last time, but there's not much of them there. No, there's not. So you don't have a lot of strength in that lid, especially with it being 3d printed.
Yes. Yeah, cuz it's a four star. It's a four start thread. And you can you can probably barely see it on the on the stream.
It's a four start thread, but it's like you get what had I don't even know what the right word for it is. But you get like single contact on all four
threads. Yeah, yeah, it's not a really good.
It's not a pressure seal. No,
it's not. Yeti never designed that seal to be a pressure seal. Right. So yeah, I'm kind of using it. Yeah,
for sure. Okay, so let's say you get this working right? You don't know the tolerance of a yeti can. So if you were to duplicate it, would it work multiple times?
That's it works on to I ordered a few koozies at different times, too. So who knows? Whatever lots they are? I don't know. It's one of the same. Amazon warehouse, right? So it's probably actually the same lot. But if this says if I can't get I know, I 100% We could make a seal. And with enough clamping force, we can reseal the can. That's possible right. Now, is it possible with this thread? I don't know yet. So the the basically if we can't get this thing tight enough with like murals, and like that, Yeti koozie thingy. What if we actually just went away from the Yeti and just 3d printed our own? You know, koozie part. I mean, it's like it's double walled. It's not as good as a vacuum, but it'll be good enough for what we I want to use it for and just put beefier threads in it.
Like it doesn't really need to be insulated very heavily. Yes. Yeah, you need to keep it you're crazy. And you don't want to be skiing down then and a beer explodes out of your check.
Yeah, in your backpack, because I froze. Oh, that is to have higher,
although you could just go skiing in really cold temperature, right? Yeah.
I think it was ever seven.
One day. Here. Ed Parker was skiing and I was like, oh, man, this must suck because I'm in Denver. And he said I was a mountain. It's got to be colder up there.
It's actually it was warmer in the mountains in Denver. Oh really wasn't Yeah, but we. So in skiing in Denver right now you get to have a reservation at the resort that you're going to the ski resort because they don't want to like overbook. How many people are going to be be there. And one of the places we were going to that morning it was two degrees. And we were like that's gonna be miserable. So we looked at the forecast and it was going to be like 14 at that resort that day as a high. We switch our reservations to a different one that was like a mountain over and It was gonna be like, in the 20s that day there and it was a difference. Yeah, one mountain over the one mountain valley over made the difference. But we looked at like what the high actually became at that one resort and it was nine. So it didn't give him get out of single digits that day. That's not fun ski. No, it's not it wasn't fun Rudel at that morning. At that one resort that we did go to for that day. It was it actually. That morning, it was still like 12 that is still not fun to to ski in. Because just cut going down actually isn't the best thing is actually going up on the lifts. Because you're just sitting there in the wind just cutting right through every seam that's in your jacket. Yeah, it's rough.
Honestly, the best days I've ever been skiing, like I will ski in a T shirt. What like you go out and it's super sunny, and it's like 60 on the mountain. It feels fantastic. I mean, everything melts and then gets all icy at night. So it kind of sucks the next day. But, but those are the best days when you're just you're almost hot out there.
So go and try out again. Next week, resealing 12 ounce cans. Version 2/3 time talking about it, part three, yeah, yeah, I got my miles out of that one.
So okay, get this recently, and by recently, I mean, a few months ago, dip trace released a new version of their software. And they offer now a tear drop function on on, on any PCB effectively. So you can define teardrops, which for those who don't know, the basically, when a trace runs into a pad on a on a PCB, most of the time, it just runs directly into the pad, and you have sharp angles wherever the trace runs. Because
the trace is going perpendicular into a a circle.
Or like a rectangular pad, you have a nice, sharp 90s. So teardrops, what it will do that is basically makes a teardrop shape, that such that the trace just before it enters into the pad opens up in a smoother angle. Yeah.
The trace becomes the width of the diameter of the annular ring. Right. And
you have all kinds of definitions like how long you want it to, like you could have a teardrop in detritus that like expands across the entire tree, like across the entire board. You have all of these characteristic that's the big
brain, your trace width is the diameter of the annular ring.
Yeah, just straight across, right. So so the benefits of teardrops, there's a few of them, the one that we're really going for is that they they potentially help with stronger pad adhesion, because you have more material, more more copper right at the pad that is glued down to the board. So this aids in rework ability. For us, that actually matters quite a bit because we do on occasion have to rework through hole pads and with non leaded solder that can be a bit of a nightmare. So you have to apply a lot of heat and a lot of heat and results in lifted pads. So teardrops kits just gives a little bit more strength to that.
Especially if you're buying on the lower end of quality of printed circuit boards as well, which won't have as high of a heat resistant glue, so to speak. That's gluing that copper down onto the substrate.
Yeah, yeah, yeah, absolutely. Which
is the wilty G rating of the boards.
And we usually get mid range TG we don't we don't we don't buy like the bottom of the barrel, but we don't need NASA grade DG. Yeah.
So another thing that is is a possible benefit of using teardrops is it prevents acid traps where these are less of an issue nowadays, but they used to be an issue in the past where if you had sharp angles in your PCBs, then when they actually etched the copper, you could get some of the acid remaining at that sharp bank.
Yeah, when they washed the acid off.
Exactly exactly. From everything I've read and everything I've experienced that's really not much of an issue anymore. So this I just gathered some benefits that have existed for teardrop
Yeah, I think I think the acid traps and modern PCB as long as you don't have like an acute angle. Yeah. In your trace. I'm pretty sure that's just copy paste of old wives tales now.
I think so. Yeah. Like if you Yeah, if you take a trace go straight out and then do like a 300 degree turn directly back.
That might be you'll get you'll get some there, but yeah, and I don't know, 90, it's not going to be an issue. Yeah, and going into a pad anymore. I think that's kind of old. Now, if you're making your own PCBs in your back yard, you're probably going to run into that issue.
Yeah, yeah, you have to pick pay a lot more attention to those kinds of things. So it's not really an issue anymore. But whatever throw we throw it out there. So the thing, the reason I brought this up is because I've actually run into a few interesting issues now that have arisen just from adding tear drops. Because as we bring new products back, or as we like, order new products, we're basically going into design files that we've built these things 100 times before, we're just adding teardrops, to just increase that re workability. But we've run into a few things that I'm going to keep my eye on, because I haven't seen any of these issues before. And now they're cropping up and fingers crossed. I don't know if you'd necessarily call them issues. Actually one of them I would, but so. Okay. There is a design practice that a lot of people do, I've been guilty of it in the past, too, that you are really not supposed to do. But most of the time doesn't cause an issue. Let's say you have
like a T SOP package, and two of the pins are just shorted in your in your design. Great example is you have an op amp that's in a buffer configuration. So the inverting terminal and the output are connected. Most of the time on offense, those are directly next to each other. So one of the easiest ways to do that on your EDA tool is to just connect the pad together inside the pad.
Yeah, go go. A direct route. Right. Crossing, we need a name for that.
shortish trace. I don't know. Yeah, he's it. Yeah, it needs to have a direct name. Yeah. It did short.
It's like a It's like going a it's like going across the pitch of the part. Yeah.
Yeah. And what's nice about that is it doesn't interfere with any other routing, you know, like, you're not going to be crossing traits in there. So it seems like a really good idea, right?
Designers to do that is like when I see that. I'm like, No, don't do that.
Yeah, okay. Well, let me ask you Why Why don't do that.
So for for me, especially since I'm in live in the contract manufacturing world is, is reducing defects during reflow. That's the and reducing false positives in like automatic inspection.
Absolutely, yeah. And, or even manual inspection, especially if you have things like black solder match boards like there or even white, they're really hard to see. And so a lot of our old products have that. Anything we design new, we don't do that, we take a trace out, you, you basically just loop it back around, you'd make two to 40 fives and bring it right back in, you can do that underneath the part, but I prefer to do it outside the part. So that said, if somebody is diagnosing one of my designs, they can just look and see, hey, these are shorted together. Well, okay, so 99% of the time, in the past, if I did the dead short trace, or on products that we have had that the the actual legs of the part don't short. In other words, they don't get a solder, jump across them in our oven. But I've noticed now when we're doing teardrops, a very small amount of teardrop tries to get applied between those pads if the pitch is wide enough, so like sy C's and some tea stops and things. It actually we've seen a much higher increase in solder paste. Actually shorting across the pads. Now electrically, that's not necessarily a problem. But from a visual inspection. That's a terrible problem, especially if something happens down the road and you have to diagnose a problem. You look at that you like, Oh, this is an issue. And then you go and spend the time wicking it off. And then you look underneath and there's a trace, right. Which is absolutely maddening because you're like, Oh, I solved it. This is the problem. No, no. So So that's an interesting side effect of having teardrops, now I I wouldn't necessarily call that like a huge issue. It's something that we need to deal with. And that would mean oh go back and you know, rev your design and just fix that. That one trace. Note no big deal. Now I ran into another thing that I've never seen before but it happened the other day. And I've made a board five times. We've went run this five times. We without teardrops. We did a new rev ran it and I and an issue cropped up where I had an op amp directly next to someone pins, I had feedback, which was just a resistor, and then directly next to those pin where that the pins of the resistor was a capacitor in the feedback loop as well. And in between those pads were teardrops now where there was just straight traces before the capacitor tombstone, but not in the long direction in the short direction it tombstone towards the resistor, and the Tomb stoning was strong enough that it actually pulled the pads off the board, which I've never seen that, wow. Yeah, it was like, and I mean, luckily, we were able to fix it, it wasn't a huge issue. And like, it's not going to be a problem. But I've never seen that before. Until I started doing teardrops. So one thing to keep in mind is run your DRC after you do your teardrops, because like your board could be flawless. And then you apply teardrops, run your DRC. Again, because you never know what your your program actually did. And you're not going to, you're probably not going to go and look at every single pad unless your board is simple enough. And look at where paste gets applied and where like the the stress is going to be applied once it goes in the oven because it applied on this component. And I think what happened is the resistor in the capacitor got placed a little bit closer together on the pick and place. One of them went in the oven kind of face first, it soldered and then yanked the other component, which, like I said, wasn't an issue before. But now the differential heating with the with the traces just cause you got a little more thermal load there. As it was as it was cooling though, because as it cooled it sealed and then yanked it off the other the other word, never seen that before. So teardrops are great, but like, it's funny, because there's these, there's there's, you know, add a new feature to the EDA tool, it's like great, it'll make my life easier. And it's and it has in many ways, but it's also like everything you do adds complexity and adds problems.
So you got to talk about a your dip trace adding a cool feature. So I don't know when my EDA tool, I say mine, eagle, the one I use, added this to the to their footprint editor. But they had this for a while and like, at least from what I know, in the PCB layout, where you could like, actually physically flip the board. I say physically again, like flip because you're virtually right, you're flipping the design over. So it's actually like you're looking at it on the backside, which is nice when you're doing like silkscreen. And if you're doing like a two sided design, the just one see the backside. So I was making a footprint. And specifically I was making the relay that I'm using on the pin heck, nothing that got Pixar. And I was looking at the end it was it's electromechanical device that relay. And guess what? It's from bottom view. Classic relay, classic relays, right? And those silly electromechanical engineers doing those layouts, those drafting.
I really wonder how many people have been bitten by that? Oh, it's like everyone, every
single PCB layout designer I know of has ran has hit that before that landmine. So I was like I was designing this footprint. And guess what? In Eagles layout, part layout tool, there's a button where you can say flip. And you can design it just how it does bottom view, and then click flip back over and it makes it all groovy in the world. You don't have to think about it anymore.
Okay, so I'm curious. Is there like a big warning? Where it's like, if you accidentally click that button, and then don't realize you clicked it?
Oh, no. It gets toggled.
Shoot, there should be like big flashing lights. It's like you're designing from the bottom? Are you really sure? Are
you really sure you want to click this button? Now I did find a really interesting bug with that button. And I think I think what it is, is that button wasn't actually designed for the layout tool. It was, well the footprint layout tool. It was designed for the PCB layout tool. But in the when you're in the library, editor, mode and Eagle, it's kind of like its own window. But it allows you to go from like the symbol, the layout, and then like the, what's called the design, I think it was called where it links those two together and you can do all the pin connecting. So if you're in the layout tool in the library editor, and you flip the board over and then you switch to the symbol It's gonna be flipped as well. But that button goes away. Oh, no, yeah, so you had sounds like, Oh, my symbol texts backwards. But that's bad. So I had to go into the layout back into the layout and the library tool, flip it back over, and then go back to the symbol. That was a really interesting like, bug I found. Which means they probably just have a, like a flag in the way it's rendering your display of like, start over and with this offset verse start at 00. Right? Yeah. Very interesting. That was oh, so I was listening to the eagle had a Autodesk I should say had a webinar about Eagle infusion of I think it was last week. And interestingly enough, you know, we always rag on Eagle and having like cream as meanie paste. So the person who was talking I can't remember his name. He didn't mention specifically why. But in the in equals code base, you can refer to every layer by its first two characters. So T, so like, TC would be top copper. Like because it'd be T copper. So all of them can be named by like the first two letters. So when you're doing like the command line stuff, you can type less. I actually didn't know that TC
is top copper would t p be top paste?
Know what is what is Z would
be top? I think it's TL
actually is no top layer. Okay, yeah, so I was wrong there. But so. So TC cream is its own. There's no other TC. But there's other there's another TP so that gives some kind of reasoning why it is that way. Is that weird? Yeah,
that's awkward.
I still think it's a translation issue, where I got translated back and forth a couple of times between German and an English because that makes more sense. But from a coding standpoint, that also makes sense. I've actually opened up Eagle right now and figuring out what else is T if it was paste was TP I'm going to fact check this person top patterns or something like that. I don't know but the thing you were thinking like pace would take priority over patterns as you'd call pattern something else.
Yeah, doesn't doesn't Eagle use patterns as something? They all have unique names for their stuff.
T p is place which is Oh, that's like the centers that's like soak not silkscreen.
Why is place
Eagle? T o which is or B O is origins and that centers? Right? Okay. Where what Where is TP? Oh, yeah, TP and BP place. So for some reason, place was used for silkscreen and so they couldn't use T paste. So it's T cream. And so, you would think they would use like T silkscreen well, they can't because there's stop mask which is T s and Bs
there's a lot of there's
there's there's like a really looking at the there's there was a cascading order going down that chain of like what was important to be named, like, quote The correct way unquote.
Oh, yeah, that's like that's like that. I can't remember the stand up comedian. He's he's got a whole bit about the the abbreviations for the states, where they're like, Oh, this is going to be easy. Alabama, L and then they're like, cool. Let's go to the next one. Alaska. Shit.
That's exactly how lazy they were like, Oh, yes. Top two yes or BS perfect. Right. Next one is pace shit
you know, what's funny is the bottom are sort of like the reverse view stuff in dip trace. I feel like there's sort of a parallel with that, where they're like, Okay, well, we need to be able to look at the bottom. So they they have a like a menu option that's just called Mirror. So doesn't matter what layer you're on. You can mirror your board. So that can be confusing, because if you don't know that you mirrored it, then it'll just always be in mirrored mode. Now, here's the thing. In mirror mode, it still keeps the text on the top layer, like, correctly readable. But on the bottom layer, everything will be flipped. So there's another function in there. That's called Flip text automatically. So if you have that engaged, and then you engage your mirror, then if you're looking at the bottom, it's like real world where you're looking, yeah, actually, you're physically flipping the board. Right? You're physically flipping the board, right? And you can you can work on the board. However, it's convenient, because it's nice when you want to actually measure things, are you doing mechanical things, but I realized there's like four worlds. There's like, the correct one where you're looking at the top. Then there's the bottom one where you're looking incorrectly. Then there's the top one where you're looking incorrectly. And then there's the bottom one where you're looking correctly. So it's, and none of these, like all of these are things where if you get yourself into them, it's up to you to get out of them. Yeah. So you could be on the bottom layer, working mirrored, and then screw things up, you know? Yep. But I guess that's all EDA tools, like you could screw it up however you want.
So this is building off our last month of discussion here. And it is an interesting article that I read a couple days ago called carmakers wake up to new pecking order as Chip crunch intensifies. Well, that's clickbait as as hell. But chip crunch. So yeah, this is a it's actually almost a direct follow up from our discussion last week, which is the auto industry shutting down by pickup truck production, from lack of electrical components. And so right now, it's like,
the chip manufacturing stop you right there. Like, isn't that funny? Like a pickup truck? Which like, is, like doesn't? Like I think of like old pickup truck that doesn't feel very electrical. It's like, oh, yeah, I'm gonna go work on Depop. Like, it's getting crippled from electronic
supply chain. Yeah. So it's, it's getting to the point where like, the automakers are pointing fingers and the chip manufacturers are pointing fingers back. It's like that spider man meme. Oh, yeah. And so basically, what's happening is like the automakers like, why aren't you ramping up? ramping up production fast enough for us? And ship makers are like, dude, y'all like, cut production, like the moment sales slump? Like, like, drop all our sales? I mean, yeah,
they run scared. Yeah.
So chip makers are basically saying, well, if you'd like, agree to like, year long contracts and actually buy the stuff, then yeah, we'll ramp it up. And automakers are like, No, we're not doing that.
It was like, when we talk with Chris Carter, why don't you just buy all this stuff? Like a year in advance? Right, exactly.
Back amazing how that works for supply chain. So for the longest time, auto, the auto industry was like, almost like the center of attention for the chip and electrical component manufacturers, like there's a whole standards on auto. What is it AQ, whatever the standard for, like resistors, automotive parts, that kind of stuff. Where you know, like, Okay, if I if I buy one of these parts out of certified, it's going to be a decent parts, most time. So, what's interesting is, what's happening now is the auto industry is being left in the dust by all the other consumer electronics and, and technology companies in terms of how much volume they're buying, and how far behind they are in the tech world. Because so the auto industry spends about $40 billion dollars a year on integrated components. Sounds like a lot. Apple spends more than that on their iPhone chips.
You're talking wait the entire auto industry, the
entire auto industry only spends? I say only it's so a lot of money. $40 billion. Okay, yeah, Apple spends more on chips just to make iPhones for their one product, one product line. So that puts it in comparison that the auto industry actually doesn't buy a lot of integrated components, because most of the time, the chips used in auto in the automotive industry are older parts and more basic components. They're not leading edge semiconductors that are going into your phones that are basically like, I mean, every single like eight months there's a new phone Knowing that you had to go buy, right? Pixel 532 Now
actually, I'm super excited because I just paid off my phone and I and my wife paid off her phone. And my bill just like went, it just plummeted. Yeah,
that's why I'm so quick. I have a pixel two that I paid off a couple of years ago. And I like refuse to buy a new phone because I'm like, this one's paid off, it works fine, as like the USB plug won't go into it.
There's too much like finger cheese and grime.
Grease, probably. Yeah, probably. So what's happening is because the auto industry is using older microcontrollers, more basic components, that kind of stuff. Chip, fabs are sort of being crunched for production. They're prioritizing higher margin parts, which are the leading edge components that like apple one, and Samsung and Sony one, like, these older microcontrollers aren't going into PlayStation fives, they're not going into the latest X, Xbox, or Nintendo or iPhone. So they get kind of just getting pushed down the assembly line because they're not high margin components anymore. It's kind of interesting way to think about it is basically like, there being up these components have been obsoleted. Not because they're bad parts or old parts. But because they're just the margins on the on them are not high enough.
Hmm, well, okay. I wonder that though, like, okay. Are they purchasing the margins or not more, but are they purchasing more parts? Are they just high runners?
I would, I would assume so. Which, which, which components? The components that go into automotive? Well, they're typically simpler components. Yeah, right. But lower end microcontrollers,
they're probably purchasing a lot more than Well,
clearly not with the only spending 40 billion a whole year.
But what I'm saying if it's 40 billion, it could be 40 billion of really cheap parts. So it would be a ton of them.
Okay, so are we saying? Well, so a lot of these components are actually not even, like, let's say they're buying a TI microcontroller? This is I'm just making this up right now, say that buying a TI microcontroller? We don't allow me to see I will use a let's come up with an IC company. Tasty chips ICs. So tasty chips ICS is got this microcontroller that Ford wants right. And for it's one of the ones that can't make F 150s For some reason right now, because they can't get the parts. And so for these, like low margin parts, a lot of times these chip manufacturers are fabbing and out to like a third party fab. Because they want their their lines they want the high margin stuff on and because they control the processes better and all that stuff. But like, oh, it's older, older chip technology, we can make that anywhere, right. So they're, they're pushing that off on these other fabs? Well those other fabs are like, what we can make like the high end parts that are going into like these graphics cards that no one can buy right now as well. And so that's basically we're just running into a problem where no one wants to build like the low margin like simple products anymore for chips.
I want to make one all the fancy FPGAs
Yes. I want some tasty chips right now. But so on semi which is one of the fabricators that is having problems with their with their supply chain right now it says third quarter this year. Is their estimate to be back on track as they scale back up. So well others I think ti said they don't have any problems because they just like banked a bunch of semiconductor material. They're fine. Tasty chips, as long as you go to the grocery store should be fine. It looks like the entire state of Texas freezes over and you can't get chips.
Oh my gosh, I've been seeing the images of the grocery stores and they're just there's nothing
it's not as bad as it was during the beginning of COVID like the beginning COVID Like there was nothing in my local grocery stores. Now. Right now it's just like, oh, you can't get like peanut butter. Yeah, it's not as bad as as
and Twinkies the stuff that last forever Oh,
bottled water was was out as well. What's the reason though? You think like beer would be old? Sold out. They had tons of beer.
What Barker's good to go. Yeah, I was good to go. liquid calories,
liquid bread.
So I got a cool op amp that I want to show off. I've been I've been doing some new components searching for a new design. And I ran into the tasty op amp tasty instruments. Tasty. That's who makes the with the tasty chip. Actually, if we ever make a semiconductor company, it needs to be tasty instruments. Absolutely, like this is done. So, so Texas Instruments has this this pretty cool op amp that I found. And I just wanted to showcase it because they label it as a general purpose op amp. And I guess it does fit that. But my definition of general purpose op amp is I guess, my my bar of excellence is way lower than theirs for general purpose because to me, I'm reading the specs on the datasheet. And I'm not just reading the marketing rank on the first page. I'm like, like going actually into it. It's like, Man, this op amp is awesome. So these are the the T O V 915. series, it comes in the 91512 and four for one, two and four package op amps. So of the the dual and the quad have are totally reasonably priced. They're about 35 cents for the dual in quantity and 52 cents for the quality and quantity. So they're a bit more expensive than you know my version of a of a general purpose op amp. But this is crazy. So let me just run through a handful of the specs on this. Because virtually every spec of this op amp is like not just like one or two times better than the op amps I normally use, it's like 10 times better than the op amps that I use. So the offset is just the typical offset is 125 micro volts, which is pretty awesome. Because like a regular tl 07 or 12 08 series op amp has like three to seven millivolts, which, if you're trying to do anything precision whatsoever like that, you're starting to get into game over territory, especially once you apply gain to that, like you're just guaranteed to have stuff all over the place. So typical is 125 micro volts on this thing. And it that increases to 780 micro volts Max across temperature, which it's not even a full millivolt across temperature and across the units, which is pretty awesome. It also has an input bias and an input offset current of only 10 pico amps, which is pretty awesome. So if you're looking for something to be good, like voltage reference, this is an excellent op amp because you have such low input bias. And only 125 micro volt offset like you can be fairly confident that your output is going to be a lot closer than using like a TLO style op amp. So you know, splurge a little bit pay a little bit more for an op amp like this. And you'll get you'll get like five six times the accuracy out of an op amp like this for a little bit more in cost. So the thing is it has a lower power rail, like the two yellows can handle 36 volts, whereas this can only handle 20 volt max and it's kind of recommended more for 16 volt. So your if your system runs on plus minus 12 volts, then Nankana work for you. But if you can run this on something like plus minus five volts, you get a ton of extra added benefits from doing this. Because this is also both input and output rail to rail op amp with some pretty awesome specs on that. So with like the full honest to God 16 volt power into a 2k load, you can swing to within 250 millivolts of the rails, which is pretty awesome. And if you use less than 16 volts of use closer to like three to five volts, you can actually get within just a few like less than 10 millivolts to the rails. So this thing starts to become way more of an ideal op amp at lower power, which if you have a low power situation that you need an op amp for this thing is actually pretty good too because it has really low power quiescent current. So it's 750 micro amps maximum across temperature for each channel of the op amp. So you could use this I mean that's not like the most amazing you can go lower than that, but in general, for a quote general purpose op amp 750 micro amps Max is not half bad for low power application. On top of that, in the single package it It has an extra pin that allows you to do shut down. So if you need an analog op amp that you can send a digital signal to and shut it down. It'll drop to like nothing. Input current, I can't remember what it is like 14 micro amps or something less than that. So you could use this also as an analog switch with a digital system on there. So all of those features to me are like, this is not general purpose, this thing is pretty awesome. If you ask me like, precision, it's got good gain. It has a 4.5 megahertz gain bandwidth product and a 20. What is it 20 volt per microsecond slew rate. So like, overall, it has like all the characters I want at a characteristic out I want out of it, it's just a little bit more expensive. And it has lower power rails, so I can't run it off of like my big power rails. But I've started using regulated lower rails for most of my stuff to reduce on heat consumption. Because I've been realizing a lot of my new designs, just, they just cook. Because I'm using I'm using op amps that run on 24 volts. And their quiescent current is, you know, in the millivolts range, well, if you start adding 20 3040 op amps in that much, you start burning watt and a half to three watts of just pure heat. So if you could just regulate your rails down and have lower quiescent current, your heat generation just goes to virtually nothing. And one of the reasons why that's always kind of been out of reach for my designs is because I don't you need good rail to rail application for that or rail to rail output. Well, this thing can get within a few millivolts of the of the rails, so I can reduce my power supply rails reduce my current consumption and therefore reduce the heat generated in myself. Because all of the new designs are not getting simpler, they need more parts. So that's sort of the genesis of how this came about, like we had a product where he's like, Okay, this thing works great. It's way too hot. So we need to find a better op amp. And luckily, I stumbled across across this.
Alright, that's, that's actually very interesting, because I might be using some op amps in that Octo temperature reader Jeep thing. So I might actually look into those two CPUs because I needed like Buffer stage. Analog Inputs.
Oh, yeah. This this is great, especially because of the low offset. Yep. Yeah. Just make sure to protect your inputs.
Yes. I want my question with these are so the ones that are multiple? Do they have a spec on like, how close they are regulated to each other or no? Is that mentioned? They're not like paired op amps in there? No, but
they do have like crosstalk certifications. I don't remember what it's I think it was negative 110 DB. Crosstalk
nothing to worry about, then.
Well, and most of your signals on on that thing are really slow. Yes. So I mean, yeah, in the in the multiples. They mean they share power, but they're going to be independent. I mean, they share the die and they share power.
Yeah, I'm assuming what they're doing is they cut, like, for the quad, they cut four out of the out of the wafer, and stick that in a package.
Well, I mean, typically they do they do a quad on a single die.
Yeah, as I'm saying a single day, but it's the same, that wafer, they're cutting that wafer into singles, doubles, and quads.
Oh, yeah, I'm not entirely sure. Maybe they manufacture singles, doubles and quads separately. But regardless, yes, we're basically saying the same. Yep. Yeah, the one thing that kind of sucks about this is, so all the big players like Mouser and DigiKey carry these, but the only people who right now distribute the T sock package is Ti Direct, which I use T sock for most of my op amp packages, mainly because all of my packages are T SOPs. So if I ever need to swap things or put new ones on there, they're easy to do. But there's, you know, the also the T SOP is the cheaper package out of all of them. But these do come in some other convenient packages too. It's just I have like a specific like all my designs use the T tops.
So I have a Intro This is kind of a deviation from when normally talk about but cyber so I've been playing a game called cyberpunk 2077 I really need to get back to it because it's like, I haven't played it in like two weeks. But um, so in that game, all the vehicles this is in the year 27 Seven alternate future. I guess. You could say an alternate future doesn't make any sense. Because the future hasn't happened yet. Regardless, one potential future one potential future in this that this game takes place in the US is that it gasoline made from genetically modified wheat. Okay, for gasoline.
So wait, wait, wait, wait. So their cars just run up here
by almost but it's actually gasoline. It's like a synthetic gasoline. So it's more close to the gasoline than alcohol.
Okay, you need to brew a beer that called cyberpunk synthetic gasoline?
Well, it's actually a better name, it's called Chu two. Because that's like the beginning of like the molecular formula for it. Okay. Okay. So earlier this week, Porsche announced that they came up with a synthetic fuel that basically replicates what tu tu does, it's probably doesn't have it's not made from like, genetically modified wheat. It's like it's synthetic gasoline. But so basically, what they did is they took, they're making this synthetic gasoline. And so gasoline usually has 30 to 40 Different components in it. And it's composed of medium ish chains of hydrocarbons. So what they made is they made a, a, a hydrocarbon fluid, I did not have a name for it, we'll call it Porsche, Porsche gas, right? This is not a gas that only has eight to 10 components in it. So it's we know they do have a name for it, they call it E, fuel, E fuel that there you go, Oh, man, that's marketing right there. So E fuel is, has eight to 10 components in it. And it's, I'm guessing, based on what they say it does. It has shorter chains of hydrocarbons. Because when you have a if you have a more pure fuel, or pure hydrocarbon that has shorter change, like the shortest that you can get as like methane, for hydrocarbon, I think it's actually is the shortest make it might be it that's like natural gas that you get for like cooking. When you burn shorter chains, you get less byproducts, you basically like the best you would get is like you get water vapor and co2. And you get less co2, the shorter the chain is as well. But so because of what they're seeing here is basically they're getting like almost no Nox, which is nitrous oxide, and which produces smog. And they're getting very few particulate matter. This is a very refined, like, genetically, you could refine gasoline into probably making something similar to this, it just wouldn't be like economically feasible. So it's easier to just make it from scratch. It's kind of why like synthetic oils are typically cleaner and longer lasting than conventional oil, a conventional oil. Anyways,
you know, interesting story. I have a buddy of mine, his father is a chemical engineer. I remember talking to him one day, because he was an engineer that worked. He had basically they had a contract firm, and synthetic oil companies would send them a request. And Royal Purple was one of the people that sent him a request to I don't know, create whatever their their oil was, and you know, based on their base, and you know how I don't know Royal Purple is heralded as as something special, right? And this guy saying like I, he's, he's a chemical engineer, he's not an automotive guy. He's like, I don't know, the performance boost. He's like, I've heard Royal Purple is what is popular. What I can tell you is Royal purples. Like hydrocarbon requirements in terms of the length of their chain are like two or three times as long as most other people that they get requests from. So there is something different there.
That's why I run I run royal purple in my vehicles. Yeah. It tends to what I've noticed with the word now, again, endo recall evidence for engine oil now for cars, but it, it holds up on viscosity longer, which makes sense on your end, if you said it has two or three times longer chains. Yep, that would mean it's holding that a longer chain will hold its viscosity longer. So going back to fuel, yeah. So what does having shorter chains and your fuel mean? It means you actually get a reduction in co2, and you get a reduction in particular, and you get reduction in NOx. And it's estimated basically Porsche saying that a running they're an engine on and this is a fuel that they can run in any currently operating gasoline. Howard engine. So it's it's definitely not an alcohol. So because like if you run you can't run a, a normal car right now on like what's called e 85, which is ethanol content 85%. And then 15% gasoline, most of the time, when you go to a gas gas station here in America, you buying e 10. So it's 10% ethanol. And there's some places you can get what's called E zero or 100% gas, which is harder to get now, you pretty much have to go into the boonies to buy that stuff. So anyways, this stuff burns in normal cars. So it's not an alcohol based gasoline, it's, they're saying you can burn in anything but running a car on this reduces the co2 output by 85%. That's ridiculous. Because the shorter chains, which actually when you do the life span, impact of the vehicle running on this verse, the lifespan impact of a electric car, it actually works out to be the same co2 footprint now, I wonder if they're including the manufacturing of the they are including the manufacturing cars, because that's that's the big thing is that's like one of the arguments people will make against electric cars like Oh, you're so putting it on the grid and running it off natural gas powered power plants and stuff like that, like, well, natural gas, fossil fuels being burned in the electrical plant, they get a much higher yield on their energy input versus output. Your internal combustion engine in your car is only getting like maybe 35% efficiency out of turning hydrocarbon into rotational energy that moves your car forward. Most of it is heat and noise. Now really cool noises but noise.
It's like somewhere around 5% Really cool noise.
Yeah. 5% of that of energy losses. Cool noises, right. So that's, that's an interesting. A very interesting. So yeah, I that's not what I'm getting out here. I'm not getting that electrical vehicles bad because they are pulling off fossil fuels off the grid. Because that's the thing is move, you can move a power plant to like the middle of nowhere, where it can like pollute a little bit and be fine. But having a city full of polluting things is not good.
If this is true, and it goes forward. I mean, there's huge impacts behind this. But this could also delay, many people moving to electric vehicles, based on like a moral or an ethical reason.
Yeah. And the thing is, is what people need to what I get with, and what I try to preach is the wrong term for it. But we'll go with that is start stop looking at like, like using fossil fuels is bad or whatever. The whole only that really matters in the end is how much energy you're using, and your system. And so the majority of vehicles energy consumption is when it's being manufactured. So keeping a car longer on the road is typically better for the environment, regardless if it's less efficient than buying a brand new car.
Oh, especially getting a new one every few years. Yeah, get a new one.
Fuse is like one of the worst things you can do. People who lease vehicles, probably is the worst for the environment. But so what we're getting at here is this synthetic fuel allows an older car to be just as basically its co2 footprint to be in less than so if you've been using this unused vehicles, now you're getting that gain on on your your energy or carbon footprint if you want to use that word going forward. Now, there's a bad downside right now. It is hella expensive to make. It is I think it's like $34 a gallon to make. But they're making it in a small factory right now. Yeah. And they're planning on ramping that up over time. But yeah, I do think what on your point is slowing the adoption? If this becomes widespread? I would say yes, we're going to have less people buying electric vehicles because of ethical moral, climate change reasons. I don't even want to fix climate change and moral reasons because they're separate things, right.
Yeah, but but they're also but they're also motivators for
yes, they are motivators. So I guess in that sense, they belong in the same sentence. But but the thing is, Is that necessarily a bad thing? Because in the end your energy if you buy car a that runs on electricity and the batteries and car B, which is running on E fuel? Well, in the end, they're both have the same carbon footprint. So is anyone actually even worse than the other one?
Yeah, I don't know like that. I mean,
so I will go towards on this this topic is, is looking at where the vehicles are used at. So city use in my in my opinion, city use I actually kind of like how, what some European countries are doing with their major cities. And either one and city centers completely banning cars. This is coming from Motorhead So, completely banning cars or making cars that get into the city electric only. Totally fine with that, actually. Like, the majority of people who use cars here in America are commuting and commuting less than 40 miles a day a day. You can get by on like a go kart. That's legit powered. Okay. Or golf carts. You don't need a f4 for sorry for you people do not need an F 150 to go to
Don't worry. They're not manufacturing them.
Manufacturing anymore. Yeah, exactly. Yeah,
I guess I guess what it boils down to is concentration of pollution.
Yes, exactly. And so I will I, I foresee in the future is for automotive, especially in how things are powered is electric cars are going to use as like the short, short runs for I don't think I don't think public public transportation in the United States is really going to take off any more than it has already. We are as a society in the United States, we're too individualistic to like get over the fact of like cramming into a bus and on a subway. Like, people have put sway in New York, that's fine. Because that is the fastest way to get around. In New York. Everywhere else. Public transportation, no matter how good it is, is also as always slower than private transportation or walking. Depending on the city, right. Man, we're getting really derailed here. I'm liking this though. Because this is stuff I think about all the time.
And you know, I realized they missed a golden opportunity here. Because they're calling it E fuel. Yeah, I was gonna say the name was the worst part about it's terrible. They could have called it new fuel. And then in the end you released gas classic, classic
so short trips, inner city stuff, electric. And then I actually really like hydrogen. That especially what type Toyota is doing the problem with
Berg really liked it to
me it's safe and cars. No problem in cars. It's you drive a truck with flammable liquid in the back already. It's no different. Oh, yeah. Yeah. It just like hydrogen just likes to make other things flammable.
It aids.
So and so I see technology like this E fuel and hydrogen. And to an extent actually, like naturally gas powered cars, as like, that's like transporting goods going in between cities, how you're running your trains, that kind of stuff. Stuff that we're I, I don't think like lithium batteries, or battery technology will get far enough to allow that dense storage, let alone chart recharging or refueling fast enough for society to accept.
If you think about how long it takes to refuel or kill a car and how much energy you are dumping into the, in the like, two minutes it takes to fill up your tank like that is an unbelievable amount of energy.
Yeah, it's people. I don't think people realize how much how energy dense hydrocarbons are. The ridiculous it's like your internal combustion engine is only converting 35% of that into like moving forward. And you're still getting like 50 miles to the gallon. Some cars are 70 Yeah. Okay. Then you extrapolate that out until like how, I'm sorry, terrible range. Electric cars are like, they're still not there yet and you still have to wait. You know, I think like to get to 80% You can do it. Like 40 minutes, maybe in a Tesla supercharger, I might be completely wrong in there. Anyways, it's not it's longer than five minutes, which is beyond the attention span of most people nowadays. What? Yeah. So I love talking about this, like the future of automotive stuff. Because I do see we need to reduce co2. This is this is one way to I think it was Brian on Twitter to skin the co2 problem. And
this is one of I mean, this seems really a too good to be true. Like there's just this article where it's like, Oh, we got this gas, that's like 50,000 times better than regular gas.
It's also 50, well, not 50 times.
And like, this is like one of the one of the biggest, if not the biggest industry in the world. And there's just this article, there's like, we'll just replace it.
There's a couple of articles, this is the first one I've actually kind of covered it. I really want to see like a paper published about this stuff like, like, the great thing about gasoline is it's has liquid good properties. And so it keeps like your valvetrain lubricated, and all that good stuff. Well, how good is this stuff at that? At doing that making sure your valve train is lubricated and all that good stuff? Is it? Does it like to absorb water like ethanol? And then that ruins spacey fuel systems? How does it eat rubber? There's a lot of other questions that you got to start asking. Now. Porsche is specifically developing this stuff, so they can run it in their older cars. Because Porsche has something like crazy number like 70% of all Porsches ever made are still on the road? That's incredible. Yeah, that's an insane number. And so they're like, Okay, we're gonna get to A, they're basically saying, okay, at a certain point, there's not going to be gas stations anymore. So how do we make sure that all these Porsches can still drive? Is there enough fuel, so I guess it's probably at least compatible enough where it's not going to, you don't have to take your classic, you know, 1960s Porsche, and tear it apart to make it E fuel compatible. So it's probably compatible enough to where you don't have to worry about that.
So one of the magic things about gasoline is, is the fact that we go, we pull it out of the earth, and then we refine it into a bazillion different things. And they all kind of rely on themselves. I mean, like, yes, one of the one of the end results of taking this crap out of the earth, and then refining it is we get fuel to move cars. But every single part of that car is also made from the crap that you pulled out of the out of the earth. Oh, like plastics, all the plastics, but all of the energy used to like, Whoa, yeah, of course, is that so like? That's one of the magical things about about hydrocarbons. Is that like, we utilize every part of it. I mean, including the co2 that goes into the environment, like we're utilizing that in a in a sense, but, but what I'm getting at is like, Okay, this synthetic stuff, if it goes to replace fuel, it doesn't necessarily go to replace all the other things that hydrocarbons.
Oh, correct. Correct. Correct. It's replacing the thing that you highly consume in your vehicles, which is the gasoline, which I actually like, the we're talking about alternate realities, because this is like cyberpunk. 2077, but like Mad Max, and they call it guzzling. Which I love that name, guys. A lien guzzling? Yeah, yeah.
But we've joked multiple times about about Parker being only one step away from Mad Max just with how many things are welded to his car. Like he does it in a very tasteful way so that's the one step away like if you don't have a nice
guy hailing clinic guitar so that's from the new Mad Max movie for those that have not watched that one yet. Spoiler alert I think that movie is like six years old or something like that now though, Fury Road Yeah, I think the running joke is like my jeep is a well be going 75 miles an hour down the freeway.
There's more will be done original Metal on you're
not it's not that bad. It's close.
You adhere to all safety. Regulations, like the whole like nothing can extend beyond your car like beyond I think it's like eight inches or something like that. All the welds go inward. So Okay, I think that's gonna have to wrap that up. Right? Yeah.
Very Yeah. Very. So. You're very interesting. I'd love to have. Okay. Last thing because I want some tips from people in Slack and probably you. So we'll pick this up for next week. But this is just to get the ball rolling is I need to this is an interesting thing. co founder, macro fab. I don't know how to start a business. Okay, I had, like, Chris church and like other people, like, do the paperwork and all that good stuff. I just like built shed. Right? Yeah. So, peloton or pinball boards. I need to start a small business that just like, does that. So I need to figure that out. And it doesn't need to be fancy. I just need to be like, Okay, people pay like that company money. And then like, I get money from it to like, build like boards and stuff with. So it's not like people paying me directly to build boards. called like a LLC or something like that.
That's probably a good route to go. Especially because it's not just you.
Yes, especially because I have contractors to pay that kind of stuff. So tips. Let me know in Slack, I'm probably going to bug like Steven next week about it. So I just wanna get the ball rolling on that one. Cool. So that was the macro fab engineering. Gasoline podcast. We're your host Parker Dolman and Steven
Craig later everyone well, thank you. Yes, you our listener for downloading our podcast. If you have a cool idea, project or topic, let Parker and I know Tweet us at macro fab at Longhorn engineer or at analog EMG or email us at podcast at macro fab.com Also, check out our Slack channel. You can find it at macro fab.com/slack
MacroFab's Misha Govshteyn and Chris Church check in with Parker and Stephen to give his take on supply chains, nearshoring and reshoring.
This week's topics are: Porsche's Synthetic Gasoline, Record Chip Manufacturing Sales for the year 2022, and the Raspberry_Pi Social Media Firestorm.
Why is there such a disconnect between component datasheet drawings and EDA footprint layout tools? Stephen and Parker dive into this on this podcast.