The US Mint Denver produces 30 million coins a day. Denes, the tooling department manager, discusses with us how production at this scale functions.
Stephen is on the hunt for the next step in his electrical engineering career and shares the shifts in the industry and what employers are looking for.
Relay manufactures hate this one simple trick that makes your “sealed” relays last longer! Except TE connectivity who has an note about this relay feature.
The Worst Timeline: A Printer Company Is Putting DRM in Paper Now
Component marking with the manufacturer in mind
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 fat engineering podcast a weekly show about all things engineering, DIY projects, manufacturing, industry news and component markings. Were your hosts electrical engineers, Parker, Dolman and Steven Craig. This is episode 323. So our first topic today is AI. This title might be a little okay, it's very clickbait. Yeah, it
was. Very click Beatty.
I do like it, though, is the worst timeline. A printer company is putting DRM in paper now. And it's some eff the was electronic. Like Electronic Frontier Foundation, I think that's correct. I'm not sure Tronic Frontier Foundation EFF. Yeah, I think that's correct. Yeah, the
product Frontier Foundation.
So what does is is Dymo, who's a manufacturer of a bunch of like label printers, has started or has been putting RFID chips in their spools that go into these label makers. So everyone knows about printer companies like Epson and brother, like for jets were like, they lock you into theirs because their their cartridges for the Jets have our have a chip in it. And so like your printer will only recognize official cartridges and have like, cartridges can only print so many pieces of paper until they run out in quotes of ink. And well, on the flip side of that, though, is like thermal printers never had that issue. Because there's no EAC, right? There's no, there's no consumable there.
Well, and the whole thing about Label Printers was you buy the printer, and then you can get whatever label just fits, and you slap it in there. And that's like one of the best parts about Label Printers.
Yeah. And so they added to Dymo added to this RFID basically lockout chip, if you want to say that into their, their spools. And the first thing I thought of though, is like, what kind of RFID chip are the using that is cheap enough to be that disposable? Right? Right, that can just go into a spool. Because a long, long time ago, Stephen and I were looking at ways to track panels on manufacturing lines, PCB panels, yeah, PCB panels, like either with stickers or stuff like that. But we kind of wanted to try like, Wouldn't it be cool if you could like, just type like, a command on like, a console or whatever. And it will tell you like, where a general location where that panel would be? was an art if you had enough RFID like transmitters and stuff around a building? You could do you could do that. But waste, we found those the RFID chips, were not cheap enough to do that. Because basically, they're disposable, because they're like, some of them do you soldered on top of the board. So you'd have like a footprint on the panel. And then some you like routed out, like put it on? Like, yeah, and they got an epoxy into the side. Those are really cool. But very labor intensive to do. And it just never worked out financially to make sense. So when they are just throwing away RFID chips in schools. So I'm guessing
you know what, real quick, fun tangent. Somebody That I Used to work with. They moved on and went to work for Lockheed Martin. And I talked to him, I don't know, a year ago, something like that. And apparently they do RFID on all of your hand tools. So if you're an assembler at a desk, your screwdriver has RFID in it. And they make sure that your screwdriver ends up in its place on your desk at the end of the day. And if it's not there, they can ask their system Where's you know, Jason's screwdriver and it can it can triangulate and find his screwdriver.
Yeah. Why did they do that?
I don't know. There must be some kind of regulations or something where they just trying to be unbelievably strict such that I think, I think it is traceability of everything. Like, yeah. How did this get put together? What exact tool put this together?
Yeah, what tool was at this desk? Yep. Oh, yeah. Okay. Now I guess if you need to track that kind of stuff, it makes loss. Yeah. Pretty cool.
But also Whoa, yeah. Whoa. Okay, label makers?
Yes. So I'm assuming. So what Dymo says is, basically, you can just put a spool in there. And now the RFID chip just tells the printer what the specifications of the spool is, which is a cool idea, except other manufacturers have done this same thing. Without electronics. I have a brother printer somewhere like in a drawer down here for like printing labels and stuff. And it uses like, it has like, pins, I guess, or like, basically switches at the bottom of the printer. And when you put the spool in, it has a grid of holes and not holes. That that either depress or not the press, this distortion is in the bottom. And that tells the printer what you loaded into it. Yeah. And that's very easy for third parties to copy. Right. Basically, what Dymo is doing is they're just locking you out into only Dymo products, because they probably have some kind of authentication, handshaking that you can't easily clone like,
well, they also say in this article, that it's it's also for auto counting of remaining labels, which fine, that's cool. But do you need all that technology to do it? Like, could you because it you know, it can it can sense that you've put a new 100 Count label in there, and then it counts every time it rotates or whatever. Give you an accurate readout. But is that worth the added cost of now paying for these roles with an RFID? in it? Yeah, no, exactly.
I don't think I've ever had a problem. Running out of out of labels like that being a problem in like a production environment. Yeah. Usually just like have a like, what we do is we just have an inspection like once a week, like look at the printers and like see how close of a reels getting empty, and we're getting close to you just put a new one in.
right not to worry about it. Yeah, no, this is not a huge problem to solve is what we're getting that
correct. Yeah. And like Zebra printers and stuff, just have a, they have a counter built in. Right. And so you just like, put the reel in, and then you reset the counter, and it knows how many it spits out?
Well, I think them saying this in their sales literature. Dymo is just kind of like getting away from the fact that they're locking you into Yeah, like
locking you in. Correct? Yeah. I do want to know what I might just actually just buy a spool and see like, or maybe someone's already hacked it to figure out what the what's actually talking between them to. Like, are they using the RFID chip as like an EEPROM to like store, how many labels are being used? So like it knows, like, oh, this spool has 643 left's. Right. And it like encodes that onto the RFID.
I think that's what the sales literature is hinting at. Yeah.
It'd be very interesting. I really want to know how, what part or what component are they using? That's that inexpensive that would work.
You know, in this article, they post a link to an NXP chip the CLR C 663. Which is a high performance multiprotocol NFC front end. Now here's the thing, I don't think that's what's in the the spools themselves. I think that that that is the chip that's probably or something similar to that is what's in the actual printer.
You see that in the article? They have
a hyperlink under in the words RFID chips. So that I don't know halfway through somewhere.
Okay. Not Found it. Yeah. Yeah, so that's CLR C 663. Let's just see what that looks like a Mauser.
I already did that as a QFN baggage. In other words, if it doesn't look like a part you would put in a real or a spool.
No, because they're like nine ollars in singles, and in quantity, they're like $7.
So I don't know, maybe whoever wrote this article might have just been like, here's a something similar, you know?
Yeah, it could be. Whoa, that is a crazy package. By look up this part number it is p n 7360 AU h n slash C 300. Why? And for those in Twitch chat, I will post this to y'all. And look at that image. That is the wackest, QFN LGA I've ever seen in my life.
Oh, that's cool. Yeah. All the all the pads are like, it's like you're going through hyperspace looking at the bottom of this chip. That's exactly what it is. Yeah. Yeah, the, some of the pads are like what you would expect. And then there's a big nice square, you know, thermal pad in the middle. But then there's a ton of pads that are all Angular. And, yeah, kind of like leg light beams in a Star Wars hyperdrive situation. So think of
a QFN 64. Except it has 128 pins. Yeah,
that's weird, actually. Okay. So it looks like the lead frame that the dye sits on before they encapsulate it in the package. It looks like they just exposed the lead frame.
Actually, I think that's what it is. Because I'm looking at page. What pages 13 of the datasheet for the part and those hyperspace leads are are just connected to the outer.
Oh, yeah. No, that's totally it. Yeah, that okay. Yeah. Okay, those look like they're pads that you wouldn't actually solder to those things correct.
The so important notes, the inner leads below the package are internally connected to the pin. Special care needs to be taken during the design so that no conductive part is present under these pins, which could cause shorts.
What? The datasheet I'm looking at which it says which could cause short cuts. Yes. Shortcuts. That's not Yeah. Shortcuts to what? Okay. Yeah. No, that's literally the exposed lead frame for the day. That's cool. Yeah, that's you don't you don't see that often? Like no, never,
ever. I've, that's the first time I've ever seen that. I'm like, wait, what? Like I just in my brain, I'm trying to think of like, how do you design that in EDA tool?
Whoa, you know, okay, so if you had to design than an EDA tool, what would suck about that is every single pad would have its own definition. You'd have to design every pad unique. Well, or quadrants, I guess. Yeah, four quadrants. But still, like that would be awful. Because, okay, in my package, dip trace. That's not a definable pad. You'd have to select each one of those pads and call it a polygon. And that define all your corners for everything. That when I
tell you not to do an equal to?
Glad they don't? I'm glad those are not heads that you're supposed to solder to.
And on the on page, oh, what is this? 96 of the datasheet. It has like the footprint layout, and it just has a big keep out area where you just can't put anything. Yeah, just don't go. Yeah. Interesting. Yeah, that's, that's not the part that's in the rules, though. Because that's that those are really expensive.
Yeah, like I was saying this, this could perhaps be similar to something that would be in the printer itself.
Yeah, the printer itself. For sure. So if you care about DRM for your label, printers, probably don't buy a Dymo 550 I think it's what it is.
Yeah, cuz they were saying like, you know, like a regular roll of labels is, you know, somewhere in the two to $5 range, and they're suggesting that these RFID tagged ones are 10 to $15 range. So your consumable part just went through the roof in terms of cost. And I don't know that's annoying. And it's also important to note that it's, it's not all of Dymo most printers that are doing this, they say like a particularly new batch.
Yes, the 550 and the five XL according to this article. Cool. Injury pristine? Well, I want to look more into like, what's actually going on in there? And what ship are they using? For the disposable end? And it's one of those things, it's like, what gets me is like we need less of this kind of stuff, right? Less less waste. This is more waste.
Will, how would it how would it necessarily add to waste?
Well, so instead of just a spool, like a cardboard spool that is Oh, recycling, okay, I see what you get. Now you have a spool that's got electronics in its
right, right. And here's the thing you've been trained. You know, you being whoever is replacing the the printer stuff, that little cardboard sleeve, you recycle it or you throw it away or whatever. You know, are you going to responsibly get rid of it now? Are you going to treat it as E waste with a little antenna in it? Maybe, maybe not? Probably not. If you've been handling it as just trash your entire the entire time you've been using it?
Yeah. So craft lab in our Twitch chat, found a link on NXP for a like the tag slash transponder components. That would this might be it. We don't know this is but it's on NXP website, which was with the other component that we were looking at. And it is the S L two s to 602 FTB x, which is a NFC RFID tag transponder. It's a three lead part and, and quantity on Mauser is 24 cents apiece, which is much more reasonable.
Yeah, it's not attendance Asia.
Quote, disposable, unquote, device.
I don't like what kind of package does it come in?
This is like a three lead.
Like a soft thing.
It's a SOT 1122.
Okay. It's like an LGA. Oh, okay. Yeah. Yep. Yeah, little pads on a package kind of thing. Like, diodes and transistors come in that.
Yeah. I don't like these packages. All.
They're annoying to work with.
Yeah. This is probably or similar to it. It looks like it works as like a built in EEPROM. And that kind of stuff.
So is it how is it powered? Are these are these spools like actually connected?
No, usually you have a antenna.
And it's just power wireless. Yeah, powered over
wireless. Okay. Usually how this kind of stuff works. Yeah, I mean, that would make sense. You know, the spool has to be within a certain distance, they probably have some tuned antenna.
See, that's why I'm wondering if they put it on on the, you know, most likely the radius of the, of the spool such that every time it rotates it, it senses again, and that's how they can count the number of labels.
Oh, no, I bet you they're just like, when you put a spool in the the RFID chip says I had 10,000 labels maybe. And then when the printer spits one out, it has a gun and that's the one that's part of how it works.
But it would be so much cooler if it was like active
or have like an accelerometer in it and it's flipped over and rotate. Yeah,
like more until as you get to pay more. It should be more intelligent.
Yeah. Could be
Yeah, I don't know. This feels a little bit slimy. But then again, printers have always been a little bit that way. You know, they you buy a $30 printer at Walmart. And then you immediately have to go out and buy $80 worth of ink.
Yeah, moment that you have to print. Your your your English report.
Yeah, your resume. Yeah, that was
my fifth thing I had at college. I actually had a dot matrix printer. course he did. And I printed all my stuff on that. And I should get another dot matrix. When I graduated I bought a little tiny laser printer for like $99 I think it was like a little inexpensive dealt one. And it's still on its original like, cartridge thing. Wow. Toner toner. That's what's called for laser toner. Right I don't print a lot though. So I should get no dot matrix though. Because those were, you know, set it up the print and
me are it noisy as hell?
Oh, yeah. But I It was cheap. Like I got it out, I think I bought it in an estate sale with like a box of paper too. And I just, I'd never had to buy paper or, or anything for it. It just worked. There, I think it ran on the printer ports. And I don't think any computer has a printer port anymore.
I had to when I was messing with my CNC about two years ago, I had to go find a motherboard that that had an integrated printer port. Because this CNC uses the printer port for all of its stepping and direction and stuff like that. Which it the software would not work over a translator over USB or anything. It had to be motherboard Direct printer port. And yeah, that's hard to find nowadays.
Yeah, I wonder. I think printers will work with those converters. It's just not seeing see, because that's you know, you're you're actually like, the bits being the bits being tickled,
it's being tickled and you you don't want any buffer in between them. You don't want any buffer in there. And I guarantee you any translator has some kind of a buffer.
Well, yeah, cereal has to be it has to have a buff. Right?
So yeah. You're not, you're seeing as he's gonna go whack if that was the case. All right. So I got a topic I want to talk about, because I kind of ran into this recently. And I've, I've sort of changed my mind a little bit on this. And this is kind of goofy. We've talked about this a handful of times in the past throughout the last handful of years, but how to mark your components on your PCB? And what are some articles
about this? Yeah, we
have we have written Yep. So So we've, we've certainly said a bunch of different things about this. But one of the things that's important, I think, with marking your components is being clear. But on top of that, like why are you even marking your components? And what is the purpose of marking your components? And yeah, obviously, you know, some some components are polarized, and they need to go a particular way. But like, what is the marking even there for? Like, who's actually looking at it?
Well, it's like, I provided a x yrs. Right? Should be good.
Yeah, should be good enough. Right? Right. So one thing that's important to note is when you okay, you give your files off to your PCB manufacturer, they buy everything. And then when it's game day, and it's time for them to build a board for you, they have to go to their pick and place and they load all the components onto reels or magazines, or feeders or whatever, they plug them into the the machine. And then they have to verify every single one of those parts. Every manufacturer I've ever visited goes one by one looks at every single part and make sure that it is correct on your board. And it's a multi step process of verification. So that would be they're checking your XY Rs, but they're also checking the actual package of the component. And they're checking your PCB. So there's multiple levels of checking going on here. And, and every every mesh PCB assembly machine I've worked with has optics on it, that will show an overlay of what the part looks like on top of an actual image of your board, which is really convenient, because then you get to see the the placement, you get to see basically where pin one is. And it depends on how the machine works. But they all indicate pin one some way. And then you can look at the board and compare that against x y Rs. And this is where it gets important, in my opinion to have your marking be extremely clear. And for a while I was against putting marking underneath components. I always liked putting marking outside of component, the landing of the component of the footprint of the footprint. Yeah, I always liked doing it outside. But I think I'm warming up to the idea that doing it on the outside. And the inside is not a bad idea. Mainly because I've run into some boards that are so densely populated, that it's become incredibly difficult to see what the marking is outside. If if there was marking on both outside and inside, I think that would that would actually solve a lot of issues, especially because if there's marking on the inside, you could basically treat that marking on the inside as you specifically talking to the machine operator, like markings on the outside or like hey, this is my indicator for anyone who's handbuilding or looking under a microscope. But the ones on the inside of the package are saying hey, machine operator, this is me telling you exactly this without having to Stand next to you or write you an email or be on the phone with you. So the marking components, there's like a gazillion different ways to do it. And I think one of the things I'm starting to dislike is marking components with a number. Like, in fact, I don't even like the idea of we've been using it a kazoo for a long time of saying pin one, that doesn't necessarily mean the correct thing. Because that your part might not have a pin one, it might have anode cathode, well, like what is pin one mean in that situation? Pin one doesn't always necessarily mean the thing that is supposed to be aligned to at the same time. So I think the verbiage around saying pin one isn't necessarily. We should say something like indicator mark or indicator pin or something because it's different
indicator I think is a good term for it. Yeah,
yeah. Yeah. And so I've run into boards where the designer really liked to put the number one next to pin one, which, okay, fine, that works, we could do it.
I'd like that a lot. If the datasheet calls a pin one, yes, for the mechanical join for.
But but here's the thing that really sucks about that. If you have a densely populated board, and you really like to put reference designators of all the components around, it's really common to have the number one around like our 101 or 11. Or think of all these other reference designators that will also have the number one, well, okay, now your PIN, one indicator for your part is now lost in a sea of other things. And at the same time, like, maybe there's like a transistor nearby that is q two, and it's really close to this pin one indicator or whatnot. It looks like q 21. Now, is it q 21? Or is one the pin one and two, like it just it gets confusing really quickly. So I would I would honestly suggest to those out there, don't use the number to represent your your directional indicator on your footprints, use a dot and and preferably use one indicator.on the outside of your package and one on the inside of your package.
So I typically do is draw bounding boxes around parts even like resistors and capacitors, mainly to show like orientation. Like if you put a bunch of let's say, oh six with threes next to each other, like let's put this input to next to each other well, do they go hot dog style? Or do they go hamburger style donuts? We put so I draw I have all my footprints out bounding boxes. Yeah. And the bounding boxes are like a, the silkscreening is basically where the keepout area line would be. So like, so you can put the parts right next to each other and basically the silkscreen just overlaps in the center. I do a I actually got been going away from internal marking. And my reasoning is silkscreen is not flat. And I've actually seen where the silk screening underneath that component, let's say q f p, this is actually what I've seen before. It was enough to cause a coplanar problem and production. Oh, really? Well, okay. Yeah. Now was that basically just extra is like the silkscreen was thicker than normal, probably was actually the real problem. But, you know, it was within tolerance of the silkscreen process, because it's not designed. It's not a super precise string is not a super precise thing on PCB. No, not at all. It's like the least controlled specification on the entire board. And so, if you have like, a QFN, you can't have any so extraordinary, that part. Right at all. And so what I generally do is put a bounding box, that's basically where the keepout area would be. And then the indicator needs to be in that bounding box. Now, I have been experimenting with putting, like the designator if he can, like a QFN you know, you have a bounding box, and then you have like, basically blank corners, right? Putting like the designator in one of the corners or something like that. Yeah. But the solution to what you were talking about though, is what's called high density design HDD and when you do the reason why HDD cost more or is because generally you don't have any. So you can't put silkscreen on right, there's no being a mess, right. And so you have an assembly document that shows pin one on all the components. My bad, I just, I just did it pin one, the indicator shows the indicator for all the components. But even let's, let's roll back a bit is your designer, putting together your footprints? There's a couple what was how long ago was this, when we talked about this, there's there's a, there's a, there's a standard out there. IPC standard for how you set up your footprints. And it's like upper left corner of the deck, that J deck standard is basically it boils down to upper left corner, or how it comes off the real the package. I always prefer coming in off the package, because that's what the machine like a pick and place is going to refer to as pin one, or the other indicates zero rotation. Zero rotation in your component is going to be the pick and place plucking the part out. And, and that's zero. Yeah. And so sure, there's like a mechanical drawing of like a footprint and most PDFs and most data sheets. But you also need to look at how it's packaged. And usually that's also detailed. Like, it'll show like a reel, and like where the pin indicator would be in relationship to the real. So that's your zero rotation. And that will help out your CM. Like, immensely. Yeah, if you just match that
well, and that's the whole point is then your X, Y, you're just trying to help your cm in this case, make things less confusing and make it easier to read it. Anything you can do to help indicate the indicator, in a clearer way is the best. I still like I haven't run into Copland arity issues with silkscreen under packages. And so I think if the package allows for it, I would still say do it if he can. You know, another thing I've run into, and I actually really liked this, but it can be troublesome. I've seen diodes, where somebody draws the diode symbol next to the diode if there's room in silkscreen, so you can just look at it. And that tells you what's entered once cathode. I've run into it however, though, where the designer messed up and drew the the silkscreen of the diode backwards, but then put the letter A anode, on an opposite pin. And so you have x, y, or s, you have the letter A somewhere, and then you have a diode symbol, and they're not agreeing, and that's extra confusing. So if you're gonna go through all the effort of putting extra stuff, just be unbelievably clear about everything and like double, triple quadruple check, then everything aligns properly. Yeah.
I don't like putting the diode symbol music's I don't have space, I just put a big, big bar where the anode should be.
The bar is the cathode. Yeah,
you're right. Yeah, we'll put the ad in the wrong spot.
Right. Right, right. And so this actually brings up a little bit of a of a side thing that I want to mention. I think it's really important, if at all possible. So you're, this is just talking to people out there who have an idea and they want to make a gizmo or whatnot. I think it's incredibly important that you own your own design, and that you be prepared to support it. And what I mean by that is if you want to develop electronics, and you want to get electronics assembled and made and then sell them as a product, be prepared to own the ability to modify those files. So in other words, you're talking with your CM and you've handed them the files, and the CM come back and said, Hey, these these part indicator markings are difficult to read. Can you go out and change the silkscreen for our next run of this like, that's a very reasonable and regular style request that you might get while working with the CM. And if you're running something like a open hardware project, where you just downloaded some Gerber's and handed it to you, cm, that would be an example of well, you don't have the ability to adjust that unless you want to go scrub Gerber's and and adjust things which that's brutal. Oh, don't do that. Done that. Yeah, we all have because we've had to, but but, you know, like, be prepared and have the correct software to be able to adjust these files. It will make your relationship with your CM a lot better. I want to make sure that I'm saying this is like, this is all like hypothetical. I'm not saying this, because this has happened to me. I just I was thinking about it earlier today, as I was looking at some some part markings and thinking like, ah, you know, I think I think we could do better. In general.
Yeah, I think if you start you start with how the CME is going to be seen where the zero degree rotation would be, and start there, where your footprint right? And then making sure that the indications clear. And then providing an assembly document,
the assembly document thing? Like, is it that's unbelievably great. If I received that I would be, you know, praiseworthy to the, to the designers, because I don't think I've ever received an assembly document that has I've had to make assembly documents, yeah, for these things. But like being given one as the CM, I don't think so.
Yeah, that's actually part of the macro crap back end, is taking all this random data and putting it into your standardization form. Yeah. For our all our network factories, right. And part of that process is building an assembly document so that we normalize all the rotations make sure the indicators are correct. By the Yeah, if you can do the indication on the inside underneath. It might be fine. I just I've seen that before of causing problems. Yeah. And so I generally don't do that.
Oh, here's a good thing. Good tip for layouts. Don't put vias right at the pin indicator? Because like, oh, yeah, most of the time that just, it just, they don't print. It just goes away. Yeah, it goes away. Like, like treat the the indicator mark is sacred and just don't.
It needs its own keep out.
Honestly, I'd be okay with that. I really would. Yeah,
that'd be interesting is a, an all your footprints is actually put a keep out on your indicator, so the silkscreen always prints clean, then that's not a bad idea. So you can't run a trace, or you can't run a via there.
Oh, also, just make sure that the indicator Mark isn't like a single dot that goes through the silkscreen, or in whatever printer they use, like in other words, like, I don't know. 10 1000s by 10 thousands.is, not an indicator, it needs to be bigger than that. Excuse me.
So designators markings, that's the biggest once you start shrinking down your designs, fitting designator markings becomes very, very difficult. It does that making sense? Like if you put a lot, a lot of resistors in a row, like sometimes, like, we start going below, oh, 603, like, oh, four twos, there's just not enough space to have a designated that's one principle. And you can fit them all by the component.
I would sacrifice designators. Before indicator marks.
Well, how do you treat that though? Do you like put all your designators in the row and then point arrows to the parts are because I've done like, a bunch of weird stuff like that. Trying to make it clear.
It depends. At WMD. We actually don't put designators on our boards. So our boards are blank, other than pertinent information.
So you really do high density design, HDD based well, even in our boards that we don't, we don't know. But that's that's that technique. When you see that a cm? Were like we do HDD or high density design. Yeah, that's what it is, is they have a process to handle boards that don't have any silkscreen on it basically.
Right? Well, I mean, we'll we'll make whatever the files say it's just when we do our own personal or the company's designs, we just do the designator less. And that ends up I mean, the thing is like when you when you start getting so low, your font size is like too, and it just ends up just being white blobs all over the place. Now in the past, I've done exactly what you You just mentioned where it's like, Okay, say I have a string of resistors. All right, the string of resistors, you know, somewhere wherever I can on the board and draw lines to them. And anything that just makes it legible. That's, that's really all that matters. So it just depends on what your company's, I guess, rules and standards are. If I have the ability to put reference designators on boards, I do that. And in fact, if I ever have boards where there's hand stuffing, or hand soldering, as much as possible, I tried to put value as well, just to make it extra clear. So designator and value, and that works with like through hole resistors, and things like that. But yeah, I mean, you sacrifice the least critical things. As you go down the list. Yeah, good enlist.
random thoughts? Yeah. What if you couldn't make the designator? Your PIN indicator?
Um, I mean, you could absolutely do that. You could. But that would also be, you'd have to make that very, very clear to the CM, because that's not a standard.
Yeah, I've never seen that before. But I'm like, a lot of times, I'll draw like a square. Yeah. For like, the indicator. And it's like, what have you put the text inverted in there? For the, for the designator? That'd be cool looking. Yeah. I've never seen that before. Maybe we can make that a standard. That'd be cool.
You know, one thing if your board would allow for it, one thing he could do, this could actually work out? Well, instead of using silkscreen for your, your pin indicator, do a little circle of exposed solder mask and just have a little copper circle or something like that, or gold circle. And then that would be like, who cares what the solder mask is? You don't have to pay attention to it. You just pay attention to those little circles.
Yeah, you could, that you could do that under a package because then you're not running into complainer problems then. Right? Exactly. You would run into solder hitting it. But generally, no. Yeah. Just don't put solder on it.
Yeah, right. You know, that's the thing you'd have to you'd have to be very explicit about those things not having paste on them.
Yeah, you have to be explicit about that. And some CMOS might see those as fiducials and then want coordinates for them. Yeah, tons of fiducials everywhere. And then you also run into the problem of what was it's hard blinking on it. But yeah, I could see a manufacturer seeing that as fiducials and wanting coordinates for them. And it's like those credentials. I remember had the first production board I ever designed. No, I think that is not the first one. The first pinball production board I designed and because we were assembling it with that what was it? It was that that old pick and place those version? No, it was before the GSM okay. It was it was a pick and place based off like at 20 extrusion Oh, very slow was mad tech or something like that?
Some Adele Summerdale that
Mattel Yeah, it was a it was a very inexpensive like hobby grade middel right pick and place. And, and the optics on it were not big enough to see because they had a qf p 100. On it. It was like a pic. 32 microcontroller. It was a QP 100 and the optics on it. Were not wide enough to see that whole package or not to see the whole package like clearly enough like there was basically too much distortion. And so to get around it, I put a fiducial in the middle of that package. Like right in the center. Yeah, and so that it could line up the that part correctly. That makes sense. It was like a point four millimeter QP which is a pretty fine pitch to assemble on that kind of machine. And so yeah, I just had a fiducial for that part right in the middle so it knew it to line up just right just just just perfect.
Just perfect. Just perfect. Whatever happened to that machine?
Um, I think we sold it. We sold it online on like eBay or something like that. Okay. Yeah, I think yeah, I think we sold it on eBay and then someone just came by and picked it up. Nice. It was a good machine. Good little machine. Yeah. Would I buy another one of those? No. I think hobby grade picking places have gotten a lot better now though. That one was a little rough. It you had to like baby it to get it to
place well, and it probably didn't. Place quickly, right.
400 places an hour. An hour? Yeah.
Oh, okay. Yeah, very fast.
Yeah, it wasn't. It was faster than hand placing for sure. Oh, for sure. Yeah. Because it didn't get tired. Right? So you could have run it for like 10 hours a day, and it'd be fine. It was mainly the babying was in like, setup, making sure all the offsets were set up, right. Because it was one of those like, it advanced the feeder by like moving the head down and like poking the lever. Right, right. Yeah, it was kind of weird.
I've seen one for like, gosh, it was like, Oh, 603 components where it would pick up and oh, 603 and drop it, put it on the board. But now the little well, for the oh three was open suit would take the nozzle, put it in there and drag the tape of
why?
Hey, it works. I guess I just like the nozzle eventually. Yeah.
And that's assuming it's like, no, I guess it would work in plastic shape too. Because if I never poked through it, it probably just like went into the pocket.
Yeah, yeah. No, it would just put the nozzle directly into the pocket. And then. Okay, and then yank it.
That's not the worst idea.
I mean, it has the it has the holes for the cogs on the side of the tape. They could have just referenced that. But
yeah, that might be better. I would say yeah. But um, that's actually not the more I think, well, it's actually not a bad idea.
I like I like I would much rather the press the little lever in advance? Yeah.
Because it didn't matter how far the lever went down as long as they went past like the index fort. Right. Right. Because it had basically a ratchet mechanism in it. Yeah. Silk Screen is. I think, I think the biggest problem was silk screening, as the primary method of showing where the pin indicator is, is there's so many different ways of showing pin indicators. And what makes sense to you might not make sense to whoever is looking at that data down the road.
Absolutely. Yeah. Also, what makes sense to you on a screen in your EDA tool? Will 100% look incredibly different through a low res camera in a pick and place machine? Yeah, that's showing you black and white, you know, not a lot of color contrast.
So my advice is always like design it how zero degrees would be your footprint. And go the extra mile. And use one of your layers for drawing like the components. So you have an assembly dock down the road. Yeah. So in your footprint, you can use like, I can't remember what it is an eagle, but there's like an assembly layer. And you can just draw your component and that it's actually really nice when you have like, an STM microcontroller that has like three dots on it. I hate and yeah, but the good thing is in those data sheets, they do show the three dots, like where they're at. Right? So you can draw them on your assembly documents.
Right? Yeah, reference this dot.
Yeah, well reference I thought, but you can draw all three dots. And so then there's only one way you can go on now.
So I did them. I manufactured a product A while back, a high volume thing. And one of the one of the interesting things was the factory that after the first board came off the line, it was validated. And then the very first thing that happened at this place, crazy, but they did it. They went every single part and they hot aired it off. And then they stuck it to a sheet of paper, basically glued that part to a sheet of paper with all the reference designators such that you had an actual like they destroyed an entire board. But you knew what every single component looked like. So even if it was like a pastor or something, you had an actual capacitor on this piece of paper with its reference designators so anytime down the road if you wanted to look at you know you had a board in hand and you'd like Alice's do We placed the right part you could go over to these sheets of paper that had all the parts ripped off of a good board and said yes that's the exact one it did they look the same, which was nuts to do for an entire board worth of parts but but they did it and yeah, I guess that's one way if you do in high volume manufacturing there's there's extra tricks like that that make problems easier down the road.
Anything you can do to make your life easier down the road it's what was it? The engineers will spend 10 hours to save repetition.
Oh, yeah, same kind of thing. This Yeah, no, this would totally be it. Yeah,
yeah. Is designed your footprint correctly the first time and then you don't have to worry about it.
Yeah. Okay, well, that was the macro fab engineering podcast. We were your hosts, Stephen Craig
and Parker Dolman Take it easy. Later everyone. Thank you. Yes, you our listener for downloading and listening to our podcast if you have a cool idea project topic, or a cool way to do pin indicators for components on your boards that we did not mention. And we would I want to hear crazier things. Like I see a lot of boards. I want to see more stuff. Especially stuff that like if you have a method that solves like everything for like, part rotations, we want to hear it. And you can let us hear it by tweeting us at McWrap at Longhorn engineer or at analog EMG or email us at the podcast at macro fab.com Or check out the Slack channel. It's backfat.com/slack Stephen and I and 670 some odd engineers hanging out in that channel. If you can show us a way to do pin indicators that solves everything. You'll blow our minds. Oh, and we live stream on Tuesdays at six o'clock central time at twitch.tv/macro Fab