MacroFab's Misha Govshteyn and Chris Church check in with Parker and Stephen to give his take on supply chains, nearshoring and reshoring.
Part shortages and obsolescence got you down? Parker and Stephen have some tips and tricks to help your design stay ahead of the End Of Life game!
Nichicon is obsoleting entire electrolytic capacitor lines. Is this a sign of how electronic component manufacturers will handle supply crunches?
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!
Oh that's a good one. I think I started last week, so go ahead and let her rip. Okay.
Welcome to the macro FEHB engineering podcast. We are your host, Stephen Craig and Parker Dolman. This is episode 260.
So, a couple months ago, I finished up my homebrew rig, right.
Yeah, both of us did this actually like half a year ago. You're right. Yeah. Cuz it was like the summer.
That was beginning of like, end of spring beginning of summer. Yeah. Was what that would be for March. April. May May. Oh, my God. COVID. Oh, it's almost wrapped. Becker. It's, it's like eight months. Oh, no. Yeah, yet? No.
Well, a few months ago is I guess that's still somewhat correct. Or just take some months ago? I did it a
few few months? No, there we go. No, because that's actually a couple's to view is
three, a few is more than two.
So like 1000 months is a few months.
I mean, now you're just getting ridiculous. You know, I have I have a huge problem with a couple meaning two, because I use a couple to mean, more than two all the time all the time.
I think. I think when you like if you ask someone, what's a couple, they'll say like two but they use in a situation where it's more than two? Oh, yeah. All the time. It's an I think couple and few is just like, it's an estimate of a low number. Usually something that you can grab in your hands. Yeah. In terms of like, you know, like bolts or like stuff like that.
Sure. A couple of bolts. I can see that doesn't just sound like give me two. Because if I wanted to, I'd say give me two. Yeah, yeah. Give me a couple of bolts. What does that mean in your mind?
Go grab like a half a handful. There you
go. Yeah. And that's almost guaranteed to be more than two. Yeah. Well, depends. So a couple of months ago, Parker finishes.
Yeah, and. And so my uncle runs a brew pub down in Alvin, Texas called it is Gordon street tavern. And they used to before COVID. And all that stuff they used to brew beer is a brew pub, and then COVID hit and they had to shut down. And there, the brewery side, of course, shut down too. And it all went to big disarray. And during Thanksgiving, I was talking to my uncle about it. And he wants to rebuild it. And then once you know COVID goes away. Once we all have vaccines. He wants to reopen the brewery side and start brewing beer again for his brewpub. And so I went over there on Saturday, and went through all his equipment. And basically, he like, tore it all down till all apart. It's all like dirty and dusty and stuff, and brought all the electronics back. And so my goal is to basically build two electrical boxes, one for pot liquor tank, which is the just the fancy word for a pot of water that you get hot. And then the boil kettle, which is that's self explanatory, at least. And then electrical box to control the boil kettle, because this system is a one barrel system. But it's the kind that you like, change the hoses around to like do the directional flow.
As opposed to your fancy valve system. Yeah,
fancy mechanical valves is not automatic yet. I don't know if I'll ever do that yet, either. I don't know yet.
I think that takes the fun out of it. Yeah,
a little bit. It definitely would like after a couple of beers and then you turn the valve the wrong way. And you shoot like half a gallon of boiling water all over the floor.
Yeah, and then it turns into sugar concrete.
Yes. Well, that before your dog starts trying to eat it all up. Anyways, so yeah, that's kind of been one of my side projects. This week is kind of building those electrical boxes and figuring out where the previous person who was doing this at his brew pub left off at and I think the idea is to basically get the this system running again and then Once it's running really well scale it up again to because they were actually getting to the point where they wanted to distribute to other bars in the area. And on a one barrel system, that's kind of impressive because that's, that's brewing a lot. And so I don't know if they are going to increase the volume, or they are going to just, again through a lot and they want some optimizations on the process or something like that. Because they have they have quite a large like, walk in for imitator. It's like seven feet by eight feet by like a least 1015
How many tanks are in it?
There is one bright tank in there, but there's a lot of imitators. Sure. So I would say at least like six, seven for mutators. So I think what they do is they probably get a lot of batches going at once. They always keep like basically the firm taters always have to be full.
Yeah, that makes sense. Yeah. Because you can, to some degree, let it just sit in there. Even if even if it's ready to go. If it's it's a few more days, it's not going to hurt anything. No. If anything, he just gets better. So this is effectively you're just building a larger scale version of what you already built.
Yeah, exactly. Just larger scale would have already built. And I think that enough parts, I've been going through all the bins that I brought, brought back from Alvin. And I think there's enough stuff to do it all minus like, I think this is like switches I need to order, but shouldn't be too bad.
Was it all done with PID loops? Just like your our systems are? Yeah, it's just a PID controller. Okay, so annual hose connections to redirect the flow, and then just PID for heat?
Yeah, yeah. I mean, I don't know if you ever went over to Eighth Wonder brewery back when Mac Feb was starting? Oh, yeah. There was one time we went over. So eighth one, back at. This was when we were at the second location Nope. Technically, to third location for Mac fab. We kind of say it was a second location, though. Because the first location was like we were rent church and I were renting a desk at like a co op thing. And we only did it for like, a month. I don't really consider that like the original location. Because, you know, we were still trying to figure out what macro fab was going to do. Right. And then we moved to a location, a warehouse that was actually right next door to Eighth Wonder brewery at the time. And this is when they were still doing like their batches and like five barrel systems or something like that. And then we moved one block down to a bigger warehouse. And Eighth Wonder took over the rest of the warehouse we were used to be in. And that's when they upgraded to like those ginormous they basically put like the biggest vessels that can put in that warehouse.
it's jam packed in there.
Are they still in that location? Or have they moved?
I'm pretty sure they're still in that location. They were definitely before COVID. And I wouldn't assume they would move during COVID Yeah, probably not. And so we were at our second warehouse location, I'll put that way. And we would go over there whenever they had electronic, electrical problems, and like and say they, they were doing like, you know, 50 barrel batches. And it's a little tiny PID controller that's controlling everything. Same electronics, just like what you're connected to is just wait there. So it's all the same.
You know, I heard something the other day. I don't think that this is true. I don't see. But but But it got me thinking like, I don't even think this is possible. But maybe it was somebody mentioned that it was either Coors Light or bud or something like that has a continuous process where I've heard that too, where basically they're they're injecting new fresh wort as it's fermenting as they're taking off fermented wort. Now, that doesn't make much sense, but that would be really cool if that actually did happen.
I mean, I could I could see that path if like think about this way if you had a long enough tube. Okay, yeah. And so you put water in one end and it took seven days to get the output technically a fermented along. And then you just have an injector where your yeast has been injected in and just
it just trickles in. Yeah, that okay, that sounds like that. That sounds like calculus one. If Parker and Stephen were to make it, remember, like they have like introduction to differential equations in calc one or Cal two. Yeah, but like The classic problem of like, you have a bucket and saltwater is coming in the top. And there's a hole in the bottom and waters coming out. Like how like at what time t is the salinity x? Or something like yeah, I would say like D Wort di t comes in and like D alcohol the D comes out. Yeah, if we did differential equations, that's what it would be.
I'm just writing out some notes. Yeah, burgers, right. Really good idea. So yeah, just building electrical boxes and then apart party go down there again and train. My uncle knows how to brew. That's not the problem is mostly training in like the process I'm going to set up for him. Like, I'm going to have all the hoses labeled. And like all the fixtures that they plug into, because it's all those tri cam. No. What's the it's the cam, cam lock cam lock? Yeah. Yeah. So they got the law years and you put it in and move the ears back and it kind of like locks onto the those things are legit. Yeah, they're also I saw how many had like he had a whole five gallon bucket full of them. And I'm like,
you know, they're actually not that expensive. Really. I thought they were Yeah, I've looked them up. They're like two or three bucks apiece. Depending on the size, you know, at the Homebrew scale, they're really not that pricey. Okay, and the cool thing is you can get them in like all like basically cam lock on one end is what you get. And then you pick the other side. So if you want like a half inch BB, or if you want, you know, whatever kind of fitting or NPT or whatever, you just choose that.
Yeah, like 10 a yen fitting. Right yeah. So if you want to quick connect your your oil line, your Jeep. Okay, um, yeah, that's it's, it should be a pretty short project. I'm hoping like, this Sunday, I'd knock out those boxes.
Well, you don't have to source all the parts, the parts of it like you're the puzzle, the guy who puts together the puzzle together
puzzle Master. Yeah. That's not as cool as like the Riddler.
So as soon as you finish those, though, you get to go and brew like a one barrel batch, right?
Yeah, he's even got all the ingredients and stuff. So it's like, it's just been sitting there.
Oh, man, I found a new brew store here in Denver, which I think is going to become my new, like, just generic brew store. Like, a lot of free stores, you go and they have grains along the walls. Yeah. And you just get to pick. They have like those, you know, the cereal dispensers. Basically,
you pull the lever, and it just shoots a little bit in. Yeah, so you have your scale. And you can kind of just fill up until you hit the your weight. So
I'm a little bit of a goober on this. But like, most of those stores, they just give you like a plastic bag and you just throw your grains in the plastic bag, put on a scale, weigh it, and then Bob's your uncle, This place doesn't have any plastic bags. And the first time I walked in there, I'm looking around, I'm like, Where? Where do I grab a bag, and they're like, Oh, so this place is literally connected to a brewery. In fact, you can walk from the brew store into the brewery and just grab a beer and walk back. So they just use all the spent grain bags and you get like the big balls you actually get the fiber glass. Yeah, you get the fiberglass so I've got this one with a with a green dinosaur on it. And like it's it's super great. You just grab a an actual grain bag. I mean, there's like the big 50 pound sacks and things. So it's just I kind of want to collect those. And you get one every time he goes. So
is that. I think they're Tyvek some kind of fibers, synthetic material,
great carbon fiber. Yeah. Yeah, I'm a goober.
I like, that's awesome. That's pretty cool.
I kind of want to like I'm building out my blue section, very, very slowly blue section of my basement. And I kind of want to use those in the decorations.
Yeah, they would be if you'd like frame them like a record. Like an album. So you actually probably buy album frames. And then because I bet that logo is no bigger than 12 by 12.
Yeah, maybe I don't know. Something like that.
That was a joke just for seeing because I was doing I was trying to do like the square but hit the frame that's in the webcam.
Something Yeah, this webcam is clearly 12 by 12.
Yes. I really need to get us recording video.
Oh, yeah, for sure. Throw it up. We Yeah, we said we were going to do that a couple of months ago.
You know the problem. The problem is I If for some reason I'm recording it on the computer, I'm also doing the recording for the audio on. It gets some weird echoes. I have no idea why it echoes from your read, which doesn't make any sense.
Yeah, that's strange. So I think I
really need to just set it up where I have we have a third connection that just comes in. And then anonymous listener, anonymous listener, yes. And then it's just recording. Now the only difference is, one of us is going to be the smaller picture. Like, in the frame, because three people instead of two, yeah, in our recording, so
just make the anonymous listener, the Mac fed logo or the podcast logo. And yeah, the
podcast logo is just the we interested in if we ever have two guests to try to do it that way, though, because squad cast only supports four connections,
does it? I thought we've had more than that. Oh, we I think we have had more than that. But multiple share a mic?
Yes. Multiple share mic. Yeah. Which makes Josh our audio engineer. Pull us air
out. Sorry, but
so. Okay, so I'm done talking about home brewery, and building electrical boxes for today.
I'm sure we will talk about it soon. Because Because rev to have no rig season. Oh, season two, I'm sorry. We're called those seasons. I forgot season two.
So season two is actually going to come for the brewery rig, my brewery rig when I move it into this room. So the side room of the side room, my garage is like my office right now. But when when my dad retires I will be able to use the office that's in the house because he won't have an office here anymore. And so I'm gonna move my office here into there. And then basically this entire side room will be I'm still gonna have my electronics in here because it makes sense. But then the other half is instead of being office is going to be brewery. Gosh, you'll
never have to leave that room. Yeah,
I'm gonna put a cut right down there.
Yeah, you do electronics turnaround suck right off the fermenter turnaround gross.
The only thing I have to figure out is if I put my brewery in here, I'm gonna take it off the cart. So it's gonna be on this like super heavy duty desk it's like man like like inch and a half plywood is doing the vent because you make a lot of steam and I got figured out how to vent it and it might be where I just open the window and put like a like instead of a window unit put a big fight 24 inch fan and then have that duct up into a big you'll make a big vent hood
Yeah Just kidding vent just go by big kitchen vent hood with a van in it.
Well, I don't want to go through the wall. That's how those are all installed.
Mainly for live it goes straight up into the attic of the garage. I guess so swampy
It's Texas. It's already swamped.
No, but I can have it I got the Vinod vent through the window. Yeah, that's the thing. Be Kind of weird doing it that way because it's like, because it would have to bend down to go into I have to figure it out. Yeah, it might have to cut a hole in the wall. So be it. Yeah. Already. Season two stuff. Yeah, I already cut a hole in the wall for an air conditioner in here. Right. Yeah, season two. I'm expecting that like probably end of summers when I'll get to it. Because I'm so kind of enjoy it don't really the only problem is just like it needs bottom drains. That's like the only thing he needs because like when you finish brewing and you just have like a gallon left in each pot, and you're like, Well, I got like a vacuum cleaner to it. Went back.
I mean, I just don't mind out in my backyard and spray them out. Yeah, but minds attached to like, your hard connected. Yeah, yeah.
So I'll probably still have hard connections except I'll just make it so that the bottom drain and not will make it a lot easier.
Gotta drill to that big bottom. That's sketchy.
Well, I gotta figure I'm gonna. I'm gonna find non tri clad pots for that. I don't know how I'm gonna do that yet. Um, I haven't used like, you know, a lot people use kegs might do that because those are single wall.
Yeah. And they make great boil kettles. Yep. Sweet.
Yep. What's your topic of
the day? Well, okay, so I kind of have some Something I ran into, and I wanted to kind of present this and see if anyone else has run into it because I have an idea for searching for components because I, so get this, I ran into this, I've run into this issue actually, I should say, not just once, but like, this happens to me a lot. And now I'm starting to think like maybe there's a way around this. Okay? Here's the scenario, you, you know, you want a part, you know what that part should basically be. And so you go to your favorite, whatever, Mouser DigiKey, blah, blah, blah, those guys, you put in all your parameters that you guaranteed know that your design has, and then it spits out, here's here's x amount of parts that that fit your filter criteria. So it's something like 150 parts, that now you have the really fun time of going through those 150 parts and seeing which magical one fit is the best for your design, right. And, and that can be either incredibly simple. If your criteria is easy. And you just at the top of the list, find one that works and you go with it. Or it could be really, really tough. If you are trying to pinch pennies, and you're trying to find the exact perfect part. Now you're having to like search every single part. So one issue that I've run into is like, Okay, I've got this huge list, how do I organize things? And how do I organize them in such a way that I know what parts I've looked at what parts I need to look at, and maybe even rank them or even say like, Oh, this one's had this one special thing. I need to remember that or, or even in a different way? Which ones do I know I just flat out don't want even though they're on my list. I just I found out I don't want?
Yeah, they're not. You don't want like, this happened to me, like three years later, in order to the same part. Can you think it's also the same part? Like it's like, it's your golden part, right? And you get it in and you're like, Oh, I remember this thing. This thing is junk.
Right? Right. Right, right. Yeah. So So one, one, okay. So let's, let's take this scenario, like, you pull up your list, you snag the first part on that list, you look at that datasheet, and you read through the datasheet and go, Okay, this is just not going to work for my design. So you go back to your search list, that item is still on your search list, and you have to mentally in your head, just know not to include it in the list, even though it's right there. My thought here is it would there be a convenient way to say like, oh, I have looked at this part, I don't want it to be part of my search anymore. Take it off the list. And take every variant of that part also off the list. So you know, like when, you know, I said there's 150 line items that show up on that search. My example from the other day, I was looking at voltage regulators and I was LDS specifically. And with with the criteria I wanted, there was a certain LM x x x part I don't remember the three numbers. But like, the first five things, were all LM x xx, just a variant of it in a different package, or tape or real bla bla, bla, and then voltage. And then yeah, down the list, there was more of that same LM blah, blah, blah, I wish there was a button that I could press and just say, Look, I've looked at the the one datasheet that references all of these LM parts. And I know none of them will work. So take all of those off the list. And that right there, I think would make searching for parts so much easier. Because I don't have a system better than the way that I've kind of discussed it here. Yeah, I think
if I don't know, if distributors like Mouser or DigiKey support it, but having like in Google, you can put a minus and then something. Yeah. And that will make sure that result that that string is not in your results. So you could do that you could do like minus Lm, and then usually u star as a wildcard, basically anything that comes after it doesn't matter. Right? And that would filter that out. I don't know if Mouser and DigiKey support that though.
I know. Okay, so I know Digi key has some support for operations like that. I know I at some point in time, it had not. Like you could say I want this but not that kind of kind of things in the search field. But I think I'm going a little bit of a different direction from what you're saying. Instead of having to re search I'm saying break down what you have already searched. So you've created a search and you have a list. The idea here is like say I click a datasheet on a part and then I find that part doesn't work because of that datasheet if there was just one button that says remove every part that is linked to this datasheet just get rid of all of those ah Ah, wouldn't automatically Yeah, it would just automatically reduce everything. And then you never have to look at that part again. And here's the thing when it comes to voltage regulators, that's not as much of an issue. But I started to think, what about connectors? Like, you know, connectors are awful at that, because you might get a family of connectors where like every pin variant is, is an individual item, if I could just say, I don't want any Amphenol xx, x or y or Sam tech, or bla bla, bla, anything connected to this datasheet take it off my list, man that would make connector certain
this would work really well for resistors and capacitor families to Yeah, because I was looking for a unusual component of resistor component that was a high voltage, like in a specific package. And you'd have to dig into the datasheet. Because actually, voltage is really something that people care too much about and resistors, like the peak voltage. So you can't really search for them on digikey Mouser. That way, you have to go all the other parameters and then go into each family's datasheet and go in bases, like are you capable of handling this high voltage and high voltage I'm seeing right now is like, over 300 volts. Yeah. So very high voltage. And, and it'd be so awesome to be like, okay, that family from the like, er j from Panasonic, I know you're not going to be able to handle.
Yeah, there's not any er j that will work out. So don't show me any er J's
Yes. And I was actually doing it by data sheet would work for most things. Now, there's some connector companies that will give you individual data sheets for something for like the families, which is sure, but that's actually a really, really good, easy way to implement it, though.
You see, I could almost see it being like, you could opt into a wizard kind of thing, where like, Okay, you do your search, you get all your results, and then you click on a datasheet. And you leave the datasheet. And maybe the search could ask you be like, Hey, I saw you open this datasheet does this part work for your design? Would you like me to save this? If not, would you like me to take it off of this entire search. And you could just say yes. And then it would remove that whole family, or like, I almost see like a part searching wizard that would walk you through finding like zeroing in on the part you want. And it would also be cool. I'm sure the big players already do some version of this, I frankly, haven't played around with it too much. But like, say I've created a search, I've done all my filtering that I know, these are the things I have to have. And then it has like a list of lists, I'll go back to my number 150 items. If I could save that list and just say, Okay, I'm going to come back to this, I know this is going to take me hours to read these data sheets, I always want to be able to look up this list, I don't want to have to remember all of the parameters I had save this list. And then I can start walking down that with the wizard until I strip away all the things that don't work. And then I'm left with two or three that I can present to the engineering manager and say like I think these parts will work. Seems really, I don't know, that seems really useful to me and way easier to narrow down things. So you don't have to just remember, Oh, I've looked at this data sheet. And this is just a variant of that. So it's not going to work for me. That'd be cool.
I think it'd be awesome. Yeah. Because Because your comment here is no more or less or expelled Excel sheets. Because I definitely have a ginormous Excel sheet that is like Parker's cool parts that I'm like saved. Unfortunately, like half of them have been obsoleted by now. But it's so cool to have the data sheets kicking around?
Well, and saving a search comes with a lot of issues because like if you save a search, and then someone accesses it later, and parts have either changed or gone obsolete or there's new parts then like how does that work? It sort of doesn't, unless like, okay, so you save the search, and then you go and bring it back up. Maybe the service could tell you oh, these things have changed since the last time you've looked at this search. That's a possibility. I don't know. They're like, it seems like there's 1000 ways to skin this cat. And all of them are compromises which Yeah, that's just engineering.
Right? What I'm saying though is is this is such a great idea because the problem with the Excel sheets is obsolete parts. And you also when you make the list, you have to put in your own description of why you put it on the list. That's how I always do it is like why am I putting this one component on this list? So I'm describing it how basically I'm hoping Parker four years from now remembers how to describe that part.
Sure, sure. You know, I'm I think also like, let's say Okay, let's see You're doing a design, you have a processor, and you need memory to go with that. And you have to go search for memory. That's not the easiest thing on earth. And there's a ton of different options that will probably all work for your situation. So how do you zero down into the memory you want, being able to, like strip things off of your search, or like Parker, you say, add notes in and say, I like this chip, because XYZ and I like that one. And then like, you know, it just helps you make decisions as you go down the line, as opposed to doing it manually, the way I've always done it, where it's like, okay, I know, I like these 10 parts, I'm putting them all in Excel, and here's URLs to the datasheet and blah, blah, blah, like, seems cumbersome when it could be easier.
Now, what we can do is you can expand this idea, and you can make it where it's publicly accessible. And that people can put in what they're using it for, like high level like, Oh, I'm using these components in automotive, or this is a consumer device, or this is a stock pedal, or, you know what, however, drill down the person wants to get for it like tags, tags, tags, or component metadata on data. Yes. And it will be awesome if you can do like popular parts, kind of like how Amazon has Amazon choice. Where like, people who search this thing has selected this item and look it
you know what, you know, it's funny, this all kind of ties into the so you want to pick apart thing or so you want to you know, like remember that that
whole idea we had where that was only a couple of months ago, right? Yeah, just a few. Right. By the way. If all right. So if anyone wants to run with these ideas, do it. Yeah, yeah. These are Stephen and I will not have time to do this stuff.
No, no, not not at all. Just remember where they came from.
Just put a little in the footer of your website to say, Hey, listen to the podcast.
Yeah, no, just Yeah, seriously, tell people to listen to that. That'll be best. So I mean, yeah, like if there was that whole, so you want to pick a relay. And then like you read about, like, here's all the things you need to know about relays. And then you go to the relay wizard, and it helps you, like drill down to the part you want. You're, you're now educated about the part and then you get to your part, but then you'd
be like, Hey, I'm designing a automotive device, and you can click automotive tag and go, Okay, people who have designed automotive things have used these relays in the past. Yeah. And these are the most popular ones. So they probably work fine.
You know, I wonder if there's, I don't know, I wouldn't know all the, like background details around this. But like, how much can Mouser or DigiKey, or any of the other big guys like steer you towards a part and be like, Hey, you should pick this one like, well, Amazon
100% does? Yeah, I guess it depends if you own the inventory, or you're just a distributor or how that works.
Yeah, like, do you have to have like a fair shake towards every single person? You know,
we because we know for sure Amazon doesn't do that? Well, so I'm shrugging my shoulders for people who can't say, who are not Stephen can't see me.
I love that you drill down into like, ferrite beads. And there's like Digi keys pick ferrite? You know,
it's just the one that they make the largest margin on?
I will maybe. I don't know, like when we keep taking ideas and then making them bigger and saying we want them.
Well, we this would be such a great tool service. Yeah. Yeah. So someone go and make that that'd be awesome. Yeah, please, I
will use it if you make it.
Or or if someone from Mac fab is listening, y'all should make it. That'd be awesome. Oh, it's
a race now. Who's gonna make it first? Yeah. So you want to make a thing to help you find a part. That's
that actually should be the title of this podcast.
Oh my gosh, we keep we keep pulling frickin band aids off of old wounded projects, you know? Yeah.
No, it's fine. It's fine. Fine. Actually, I have a solution to that. I say me, Chris Carter, who's been on the podcast before from Mercury Inc. He gave me this really good idea that I want to implement for the Slack channel. I'm not gonna go any deeper because I want to actually do it before we actually like, say anything more about it. Like, that's what I'm trying to do with my projects. Now it's like, is I want to at least get like halfway there before I even say anything about it now, right? Like with the my uncle's breweries like, I already have all the parts here. And I have to do it. But I guess technically I don't have to do it but being a good nephew. Is that the correct
If he's your uncle your his nephew. I think that's correct. Yeah. We have the same last name goddamnit.
Actually, I came up with an idea for a project. And not not a project for Parker or me but a project for our listeners actually, the other day, and I'm not gonna say anything about it. I'm not saying anything about it. The whole real whole reason I even said that is because Parker is like, these are the criteria that must be met before we tell our listeners what that is. And you know what it? It likely will never happen. It may. But it likely won't. I'm just bringing that up because like Parker's just being a gatekeeper now, and I appreciate that. I think it's,
I think it's an idea that we have that we know we're never going to do like this. This part, picker wizard. Hey, who's gay blah thing? I know we wanted you wanted to do the so you want to pick up blank thing. But that never happened. So I think at this point, we're cutting the reins on that one.
I mean, it. I hope it happens one day.
Okay, so yeah, we're cutting the reins on that one. So yeah, but so if it's a project that you we want to do, I think we have to wait until at least we've done like 50% of the work. Okay, that's fair. You have to be able to Okay, how about this, you have to be able to take a picture of something physical.
Oh, so So it's basically Kickstarter rules now. Yeah, Kickstarter rules.
I think actually, screenshots of schematics and stuff is enough to my, in my opinion, sure. Something has to actually exist, that you can like, show to someone. Got it. So it's not even Kickstarter rules. It's like Indiegogo rules.
Yeah, and in general, unless guess unless we say something about it. What we say on this podcast, like, like a lot of these ideas, you can just run with them. You know, somebody wants to, especially the idea tank podcast. If you have ideas about or if you want to implement those. We should get those guys back on. The pilots are so much fun. I love that was such a great podcast. Last one. We did some fun.
What episodes were those?
I don't know. One was in the bomb shelter. And one was
at the pre heart. Yeah, a couple months ago, right? Yeah. Yeah, one was map episode 77. Because that was a couple months before Hurricane Harvey, right? Yeah. Because hurricane Harvey was August 2017. Yeah, I think that's correct. Yeah. And that was July 2017. So it was seriously a couple of months before Harvey and then the other one was met Episode 223, which was a couple months ago.
May 6, that really wasn't that long ago.
It's almost wrapped back around now. And that was the inverted success funnel. Which was the premier pyramid scheme that was inverted
That was you. I think you had one of the best ideas on that last one. Because you had the go away. Ai
Yeah, that one was fun.
Everyone go back if y'all haven't if you're new on the podcast, go back and listen to those because they are amazing. They were that was fun. Okay so serious news now seriously,
actually real quick if we ever do another another idea tank podcast, we should have our Slack channel come up with an idea like a collective idea and then present the Slack channels collective. Id
like we have to sell the Slack channels ideally.
Yeah, it's like Shark Tank where we have to sell slacks idea to to the other guys. That
could actually be because right so the format of the idea tank podcast when they're on our podcast is we just come up with some random ideas and that's about it. I liked the idea of pushing it one step further is you have to sell your idea or product to the other people in the in the podcast.
Yeah, I like that.
So you have to come up with like, $1 amount and all that good stuff. Yeah, that'd be awesome. Okay, basically Shark Tank the podcast, right? Right. Yeah. Okay. serious news. We're 40 minutes into this thing. serious news time finally. Okay. Finally. Okay, so chip shortages are happening. We talked about this a couple of weeks ago. But this is more industrial. And industry news is shipped shortages hit even as auto chip sales climb. So last quarter, quarter for 2020. Actually, the whole year 2020 was not very good for automotive sales. Not a lot people were driving, people were not buying cars. But for some reason, in quarter, four people were like, I'm sick of not buying a car. So I'm gonna go buy a car. I have no idea why consumers would do that. And quarter four last year. That's what happened. And the supply chain is so stressed on the electrical components that go into cars that Ford and Fiat Chrysler are shut temporarily closing factories, because they do not have enough components to build their electronics that go into their cars.
That's bananas. Is that like, across the entire like, like their entire fleet? Or there's only a couple
of factories are shutting down. Okay. It's not like the whole fat whole thing sitting down. Yeah. But yeah, they're shutting down. I think it's a couple of their lines, just because they cannot get enough components. Wow. So I was thinking about this. This, that was an article from fierce electronics. So I was thinking about this. And I'm like, What else has gone bonkers in the automotive? Because, like, recently, so Apple, it was a couple months ago, announced they wanted to make self driving electric cars by 2024, which is, at this point three years from now.
Which is crazy to think about. And so it's like, okay, a, your average lead time for like some of the really in demand, automotive grade components like MOSFETs. Now using you're using to drive your motor controllers, for your big electric motors to move the car. Those are having like some, some of those components have lead times of like 36 to like 100 weeks. So basically, it's like half a year to two years of lead time on these components. And I'm thinking, Okay, you have this player, apple that's coming into the space that wants to hit 2020 for the problem under leverage, like third party manufacturers, all that good stuff to build these cars. But you're talking, they're buying components now. To hit that 2024. So you got this new player, that's gobbling up all these components. And then you also have Tesla, who has been a long term player in electric cars. But in 2020, they hit their highest sales ever with basically half a million cars sold, which is 36% increase from 2019, which is ridiculous to think about. So they have a 30 36% increase, which means they are using 36% more parts minimal. And then there was a VLSI research came out with a report where basically the fourth quarter automotive chip sales, so like the, the components that go into electronics for cars, was 30%. Greater in quarter four, then previous quarters,
people in saving money all year and they're ready to blow it on a car.
Exactly. But yeah, the automotive chip sales was $6.2 billion in three months at the end of 2020. Wow.
Made up for the rest of the year.
Yep. Well, almost. Overall, it was 6.4% down over the whole year compared to 2019. But that last quarter, q4 was 11% over from quarter four from 2019, which is ridiculous. Basically, it's like the greatest it's ever been after being down for three quarters. Crazy, almost made up for Yeah, the whole year. Almost it so it's one of those like, you got all these big manufacturers and I see manufacturers are just either reluctant to scale up or just they can't scale it fast enough. They can't build factories fast enough. I don't know what it is. Because we saw this back in, oh, man, was it 17 or 18? There was a bunch of icy like, it was actually like sea moss level components were were really hard to come by for a bit. It was the factories didn't want to scale up. And and then I think it was like a year or a couple months after that. It was chip components, didn't they? Were you remember all the capacitors shortages? Oh, yeah. And they, basically they, they didn't want to scale up any of their lines. And they just basically redid all their offerings, I guess. Component lists, I guess. It basically just reduced how many SKUs? They had. Yeah, big consolidation. Yeah, they just consolidated a lot of the component lines. That's how they solved the supply chain issue. obsoleted, like, half the components on the market?
Yeah, I we we don't do this anymore. But it was like a one time like roll of the dice. And we bought like a million capacitors or something like that. And they were all bad. Every single one. It didn't actually end up. It wasn't that pricey or anything. So we didn't lose a lot of money. It was like a
$80 reel on Alibaba.
One real with a million parts.
Just like a tractor wheel. Yeah, I
was six feet wide. Ya know? So it was just like, oh, well, we rolled the dice. I say roll the dice because we gray marketed that. Yeah, we don't. We don't have our gray market stuff. But it was like, let's just try it because of shortage shortage.
And I was at there. No, that must have been right when you went to WMD. Yeah, it was
it was the first couple of months. They handed me reels the capacitors be like, can you check these these seem bad? Yeah. Yeah, they're bad. I remember
one of my friends bought a it was during that whole shortage to bought a reel of sketchy components off off. eBay off sketching. opponents.com
Yeah. Oh, yeah. eBay, right. Yeah, eBay.
And the first couple of the pulled off the reel were fine. Yeah. So you had a manufacturing run done of them. And about the first 100 Were okay, and then the rest were all dead shorts. And they're, they're supposed to be 8.1 micro farad. Capacitor sucks. So whoever loaded whoever counterfeited the real but good components first, so you would test them right? And it'd be like, oh, yeah, these are finest and what do you noggin right there? Yep. No returns on us components. Even if they're dead shorts. Don't buy from unauthorized dealers. Kids
know you roll the dice every single time and you will get bit you always always. Yeah, eventually is what I mean. You may get lucky but you will get bad eventually.
I remember imagining like it it's the world it's 2017 there's a capacitor shortage a van rolls up next to the engineering complex it says on the side and spray paint free capacitors Yeah, guy with a trench coat
opens it up and there's reels he kid you want some point ones I got cabinets
you got some geeky cons some fine gold.
And they're not even gold. They're just like yellow spray painted. Yeah, the dip dip. Dip electrolytic Oh man. So one other one other quick thing. I've I've created a new style of circuit building. So you've created a monster is what you have no no no. So So we've actually created one already on on the fab and I have leveled it up I think and created a new one so so we you know there's surface mount there's through hole. The next level that is like of sort of insanity is Manhattan style building. Go google that and and just look at that like beauty that people make. And then and then with with our buddy Roz on a podcast we came up with thunder ohms style, which well, you got dead bugging. Oh, I apologize. Yeah, there's dead bugging above that which dead bugging is effectively you flip a chip upside down, glue it to the board and then just wire from pads to whatever the pins are. Most of the time you dead bug because you got a footprint wrong or you use a connector. I'm not wrong pin up. So uh yeah. And then in the level up from that you got thunder, Rome style, and thunder own style is Mad Max anything goes in a way. But there's still like some rhyme or reason a little bit to it a little bit. It's ugly, but it gets the job done. So today I am going to declare that I officially created the next level and I we're calling it Cthulhu style. Manufacturing, it's it's point to point surface mount with any any PCB connection done with a flying wire. So I posted up a circuit that I had to build today. It's a testing a new, a new concept. And basically, I was looking at all the parts, I was like, this is going to be a nightmare to slice up this board and cut traces and like bend legs and stuff. I'm just going to cut through loose style, which is basically you take the part you solder surface mount components to the legs where you need them, and then just wire it all up. And and this thing was was it's three dimensional in terms of like I have, I have jumpers going underneath parts and like surface mount parts hanging off the Flying Spaghetti Monster. Yeah, basically, basically, it's like all of your nightmares come alive and go through Lucite style manufacturing,
I think as HP Lovecraft style Yeah, love.
And, and so so so the story of this, this circuit goes like this, we use voltage references for a lot of our analog circuits at work basically, a really nice stable voltage that we can have the rest of the circuit be controlled by most of the time, it's like five volts or something like that. So our potentiometers we we know the range, they go from zero to five, or they go from negative five to positive five. And then we can design around that. If you don't have stable references, then what is your circuit, right? Like, it could be anything, it floats all over the place. So a few or a a couple of months ago, I designed a new reference circuit that uses a tilde 052 op amp, which is a pretty nice little precision op amp, that's not too expensive. And in conjunction with a reference voltage generator, the MCP 1501 t, we can get virtually anything we want from all of our analog references to our DAC references to all kinds of stuff. So those two parts together give us virtually everything we want. And we've already used this reference design on a handful of a handful of our circuits. So we we get down to throwing this on to a new product or a it's not a new product is an old product that we've done a board rev on, and we wanted to throw this reference on it, we get the reference and start testing the new, the new prototypes. And like you turn a knob and the reference just floats all over the place. It's doing non referencing things. And what this kind of boils down to is perhaps a little bit of marketing WINC in the datasheet, but also not fully reading every aspect of the datasheet from A to 052. Because the TEALS 052. If you look at the datasheet and actually claims a pretty high output current, and like it can drive, moderate load. But if you start digging deeper and look at the charts, then they're like, yeah, it can drive. Okay, loads, but its output voltage, just dive bombs to nothing. So like if you're, if you have its output voltage low, you can drive more. Yeah, it claims that its absolute maximum output per channel from an SOC a op amp can drive 80 milliamps. But like, maybe at like point one volt, or something
like it's, it's a reference that you're not going to use Yeah,
like it's the absolute maximum make it claimed as 160 milliamps for the whole IC, which is incredible. I mean, like it's ridiculous. But in reality, it can actually only drive, excuse me, like a regular offense, it can drive 10 to 20 milliamps before you start seeing some degradation. And you wouldn't necessarily know that unless you looked at the charts that are that are down there. And I may have missed those. In reference to this new circuit, like in our older circuits is a reference voltage. Typically you're not demanding much current from a reference voltage because you want it to be really stable and Most of the references are going to be high impedance things that are just looking at the voltage effectively.
So are you sure this datasheet just didn't buy you a really fancy lunch?
I may have I, you know, I that was a couple of months ago, I don't remember.
So, salesperson joke,
that's just a thing. Yeah, you gotta watch out for like, I don't know, like this, this op amp, I designed it in as a higher current than it then than I thought it would be. I didn't fully check that. But like I said, I designed it for some of these other previous products that we've done. And it works beautifully on all of those, it's just, we put it in a very high current demanding situation that didn't check that. So it's different criteria than the original design. So regardless, this whole circuit kind of Craps the bed, when you demand more juice out of it is what the moral of the story is. So I needed a solution to that. And instead of like finding a new op amp, and doing all these other things, I just decided to go with an op amp transistor buffer circuit, which we have a link to an all about circuits thing with it, I've used this circuit multiple times, and it makes a great little current increaser basically, you can, you can demand more out of your, basically you use the reference voltage to go into an op amp buffer. And the the op amp buffer itself controls a transistor where the transistor is doing the heavy lifting, of dropping voltage and source and current. So between those three, like I just basically had to build this Cthulhu style circuit up in the air and and run that and you know what's great? Like, it works freaking beautifully, like, we're good to go. Basically, I just need to pick the right transistor for production.
So how much error does that buffer circuit add to your reference?
So my reference is, so my reference is three volts. And it is a point 1% initial accuracy, then I put it in that to 052, which has really low offset not really low, but but it has moderately low offset. And it's it's gain error is good. And I measured, I measured the output at close to point 1% air across the whole day, basically, I gain up that three volts to a plus and minus five volts. And I was well within five millivolts of hitting that target. So pretty good. Let's put it that way. And that's obviously at room temperature, you know, if you if you sweep the entire range, it wouldn't be as nice. But if I'm getting an initial accuracy on both of my reference rails of within five millivolts of five volts, I'm happy with that. It does take some precision resistors to do it, though. Basically, basically a circuit like this is as precise as you want it to be. You just how much money do you want to throw at it being precise?
Yeah, how much? What's your tolerance stack up? Basically, at the end.
The good thing about this though, this particular circuit, I swept the load from its quiescent just basic, I think it was like 10 milliamps, and I swept it to its maximum voltage load, which was 40 milliamps, and I got virtually no change in the reference voltage, which is what you want, you just you want it to be linear, well, not just linear, you want it to not respond to changes in load. And that's exactly what I was getting with this.
And looking at the circuit now, because I've never dealt with these kinds of circuits before. And what I'm noticing is there's no capacitors in it, in terms of like in the feedback loop and that kind of stuff. Which is good, because precision capacitors are very expensive, which as compared to a precision resistor, which is not as expensive. Well, okay,
before before you go too much further with that. In the real world. I do add capacitors to the feedback of the circuit. And it's they're small, typically somewhere in the range of like 100, Pico farad. And that's basically to keep the phase margin of everything like playing nice, such that it doesn't it like it aids instability, and just reducing the ringing on your feedback loop. Well, I mean, it's a reference voltage, there's no there's should be no step changes should be no ringing whatsoever. But just making sure because Because instead of the op amp, like legitimately feeding itself back and now has to control the transistor that feeds back to it's with the inverting pin. And so, at the time when you first power things up and it comes up that capacitor helps kind of slow things down so it doesn't freak out and go nuts And the output of the op amp doesn't just saturate. I mean it, there's a chance that it still could, but the capacitor there helps make sure that it's under control. Let's put it that way. So, anytime, anytime you can add a little bit of extra control and swap that out I typically try to in fact with most of my designs, I will add pads for capacitors and feedback in op amp feedbacks and just make them do not installs on the prototyping in case like I do find some instability I can just slap a cap in. So yeah, the circuit in this link that we have here is like the bare minimum like the breadboard version of of it. But yeah, adding adding a small capacitor in the feedback loop is probably not a bad idea.
Sounds good to me. Cool. So that was the macro engineering podcast we're your host
Stephen Greg. Lane everyone take it easy
Thank you, yes, you our listener for downloading and listening to our podcast. If you have a cool idea, project or topic, let Steven and I know or if you take one of our ideas and run with it and make a billion dollars. Remember and listen to our podcast. Tweet us at Mac fab at Longhorn engineer and like EMG or email us at email@example.com Also check out our Slack channel. You can find it at macro fab.com/slack
Nichicon is obsoleting entire electrolytic capacitor lines. Is this a sign of how electronic component manufacturers will handle supply crunches?
MacroFab's Misha Govshteyn and Chris Church check in with Parker and Stephen to give his take on supply chains, nearshoring and reshoring.
Part shortages and obsolescence got you down? Parker and Stephen have some tips and tricks to help your design stay ahead of the End Of Life game!