#13 – Chips And Fabs And Garages

Chris Gammell: Welcome to the Amp Hour. I'm Chris Gammell from Chris Gammell's Analog Life. How are you doing today, Dave?

Dave Jones: This isn't Dave. What? Where's Dave? It's not Dave. I think you must have called the wrong guy this week. Oh my god, who did I call? It's Jeff Kaiser again, Mighty Ohm. Oh, hey Jeff. How are you doing, man? Hey, pretty good. How are you doing, Chris?

Chris Gammell: This works out well because Dave's actually on vacation and he can't talk anyway, so what a weird coincidence.

Dave Jones: Yeah, that's really, really strange. Well, it's good to be here again. I guess, you know, no notice is better than no notice.

Chris Gammell: Yeah. Yeah, that's right. So, well, thanks for coming back. I think we had really good response last time you were here.

Dave Jones: Yeah, well, I'm happy to hear from you again. I guess it means that the last show didn't go terribly wrong. No, it went pretty well, actually.

Chris Gammell: I don't want to talk about the Uno again. I mean, we talked about it. Oh, no, no, no. I think we beat that one to death.

Dave Jones: Yeah, and I think I talked all I ever need to talk about the Arduino Uno ever. I mean, I just said it there. That's probably enough.

Chris Gammell: Oh, God. Yeah, what are you getting it for the word? You know, like Massimo Banzi is throwing you a couple bucks every time you say it. So where's Dave? Dave? You know, I think he's in the middle of, like, Forest in California.

Dave Jones: Oh, I think he's in the Sequoia National Park.

Chris Gammell: That's right. Yeah, he had that on Twitter today or the other day.

Dave Jones: Yeah, I saw that. So he's not around. Yeah, you know, actually, one of the things I saw that he tweeted about, if I can interrupt for just a second. No problem, man. And you can stop me if this goes off on a tangent, but I saw that he has been all over California. Yeah, yeah. And it looked like, I guess he was there for a Renaissance conference, right? Some kind of a seminar or something. And I didn't get a chance to watch his videos on that, but I did watch his Twitter stream. And I saw he went to one of my favorite places in the Los Angeles area, which is Apex Electronics.

Chris Gammell: That's right, because you used to live around there, right? Or you lived in San Fran.

Dave Jones: I used to live in San Francisco, but before then I lived in San Diego, which is not too far away. But I have a good friend that lives in the Los Angeles area, and I actually went down, I guess it was about a year and a half ago, with the express intention of visiting some of the really cool electronic shops that are in the Los Angeles area.

Chris Gammell: That's crazy, because I would not have thought that L.A. would have electronic. I mean, like, you think L.A., you do not think electronic shops.

Dave Jones: L.A. has everything. If you're into do-it-yourself anything, I can pretty much guarantee that L.A.'s got it. I mean, there's stuff for people that do woodworking. There's stuff for if you do metal shop, you can buy surplus materials. And this friend that I've got down there is into the electronic scene, and I've known him for a long time. And so he knew all the places to go that were in Los Angeles. And so we actually went on a weekend, and I flew in. And this is whenever I was working on that Wi-Fi radio project that I mentioned last time. Yeah, so he did the woodwork and did the really nice case. And so we actually went to, like, a hardwood veneer shop and bought the veneer. And that was really awesome. But the best part was we got to go to some of these electronics surplus shops, which, as I think you know, the Bay Area has quite a few. Right. There are still a few left. There used to be far more. And these are places where you basically can buy electronics junk, most of it for the price of the metal. So really a deep discount. And so one of these places is called Apex Electronics. And I can't remember exactly where it is. It's somewhere north of L.A. And it's the kind of place where you can buy used vacuum tubes and wire and meters. But then they also have weird stuff like bomb casings. And I've got a couple pictures on my site of whenever I went there. And hopefully Dave took some video. He said he did, yeah.

Chris Gammell: Awesome. I think he mentioned that it was going to be like the Akibara tour. I think I'm saying that. I hope I'm saying that right. I don't know if I am.

Dave Jones: I don't think it's quite the same thing. It's not a city devoted to electronics. It's more like a junkyard. No, that's actually exactly what it is. It's a junkyard devoted to electronics components. It's really amazing.

Chris Gammell: L.A. wants your crap. That's right.

Dave Jones: Yeah, and, you know, actually, the interesting thing about it is that I think the reason that Apex stays in business, and I could be totally wrong about this, so someone please correct me, but if anybody knows the Apex guys, I think one of their biggest customers is Hollywood. Oh, yeah. Because all the stuff that's there, and if you look at the pictures online on the Apex electronics site or if you look at Dave's video, if he posts it, you'll see that everything just wants to be a prop in a movie or something. It's just you can get all sorts of mad scientist kind of stuff and all sorts of neat things, and I think it's really cool, so I'm definitely looking forward to seeing what Dave recorded there.

Chris Gammell: Yeah, that's great. I imagine, like, maybe he'll be able to find, like, you know how they, in Back to the Future 3, you know, he's a big Back to the Future fan, how they replaced the microchip with, like, all those vacuum tubes on top of the DeLorean. Yeah, yeah. That's what I imagine when you say, like, the props that you would get there. No, absolutely. Because it was a tube base and everything.

Dave Jones: That's exactly what this place is, and there's actually some other places like it, but I don't know the names of them. Yeah. I know the equivalent in Silicon Valley, which is a weird stuff warehouse in Santa Clara, or no, Sunnyvale, I guess.

Chris Gammell: Uh-huh.

Dave Jones: The same kind of place. Like, you can wander there for hours just looking at weird kind of semi-useless junk that, you know, to the right person, I guess it's worth a lot. Yeah. Really cool stuff. That's awesome.

Chris Gammell: Well, let's move on to the shout-outs. We don't have any of this week, but one that I wanted to give a shout-out to is ChipHacker. So you said you didn't know what ChipHacker.com was.

Dave Jones: I don't think I've ever been there.

Chris Gammell: So it used to be ChipHacker.com, and recently there was ChipHacker and another site, and I think Electronics Exchange, perhaps? But basically they merged, and now they're... So you know Stack Overflow, that site for programmers? I'm not familiar with it. No. Okay. So basically Joel Sposky and I think that's how you say his name, and a couple other people, they started up a new type of forum, and it's this really great way of doing it where it's kind of merit-based. So the more you post, the more rights and privileges you get. The more people you get there, the more expertise, obviously. That's any board, right? But I really like the format because it's kind of like combining a wiki. It's combining a message board. It's combining like a Reddit-style voting system. And it's just this really awesome, awesome way to get information. So anyone out there listening, I would highly... If you're not already looking at it, it used to be ChipHacker.com. If you still go to ChipHacker.com, that'll work and take you there. I'm not sure how much longer they'll maintain that. But now it's electronics.stackexchange.com.

Dave Jones: Yeah, I'm actually... I've just got it open right now. And yeah, it's kind of a neat idea because the people that have the most interesting questions and projects get voted up in the list. And so you kind of get to see really cool things. And it's not just one person posting, but it's like a discussion about what they're working on.

Chris Gammell: Huh, yeah. And I've used it recently for personal projects. And the really nice thing about like a stack... I think they're called... I think Stack Exchange is the actual open source software that runs the actual message board. It used to be called... I think the Stack Overflow is the one that was the initial one. But then once they developed it enough, they actually opened it up to the community. And now you can download it and start your own board. Kind of like a Wikipedia. Oh, I totally forgot where I was going with that. Yeah, either... Oh, that's what it was. Basically, it's really nice because there's a lot of mechanisms in it to have a really broad range of experience. So, you know, a lot of people that are beginners would often... You know, you don't want to... You either got to go through and open up everything and try and find your answer, which is always a good idea. You know, you're going to get in trouble if you don't actually try and search for it. But the thing is, even with beginners, they're very much encouraged for the easy questions to be asked there as well once you've searched. Whereas on a forum, that might be a little bit more looked down upon. You know, there's always different levels of forums. You know what I mean?

Dave Jones: Oh, yeah. You got to watch out for those forums where if you're the noob, you get... Yeah, exactly.

Chris Gammell: Yeah, yeah. So, this plays really nicely for having that broad range of experience where you'll have people that are... And because of the voting system, it'll, you know, bring the more experienced answers to the top. And if you look at it, you're always going to see the most experienced answers at the top. I mean, so, it's very nice in that regard. And yet, you can still add in... You can comment on answers. And you can comment on the question even. Like, oh, this shouldn't be here. Or you can... If it's a community wiki, it can open up and you can actually modify it if you have the privileges. And much like Foursquare or a lot of other people are doing this now, they have badges too. So, if you... You know, it's really interesting psychologically. You know, it doesn't seem like that would really do much for you. You know, it's like if you get a badge, who cares, right? But at the same time, it might start motivating you towards certain behaviors. So, if you're supposed to... You know, once you get the privilege to edit, if you actually go and edit something for the better, you might get a badge for that. You know? And so, it actually encourages involvement. Like, and there's like different levels and blah, blah, blah. Not necessarily... You wouldn't go there just for that. You go there for the electronics knowledge. But it's very good. And if people out there listening have never been there, once again, it's electronics.stackexchange.com or chiphacker.com for the time being. And I like it a lot. And the more people that are there, just like any message board, you know, it's... Yeah. So, I love it.

Dave Jones: Yeah, very, very cool. I'm just actually... I'm reading through it right now. And there's something for everybody because they've got a lot of questions or topics about microcontrollers. There's soldering, surface mount electronics. Somebody's asking what the cheapest TQFP44s and SO28s are to practice SMD soldering. And there's people telling them where to get cheap parts. So, it's cool. It's like any question, no matter how obscure, has some answers. And it seems like people are being genuinely pretty helpful, which, you know, in forums, sometimes you don't get that as much. Yeah. Yeah, this looks really cool. I'm definitely going to check this out.

Chris Gammell: And, you know, I've asked the question before in my blog about, you know, where do you really go for some of this stuff? Because a lot of times it's not out there, you know, or it feels like it's not out there, or it's fragmented. You know, like there's, you know, like Dave's site, EEV blog, they have forums. Adafruit has forums. I think your site has forums?

Dave Jones: Yeah, I've got forums, but they're fairly small just because they're only to support individual projects.

Chris Gammell: Right. And, you know, it's nice because the software that's out there is very, you know, people know how to use it. So, that's good. So, that's kind of a downside for this new Stack Exchange thing. But, at the same time, I think, you know, if there actually was a centralized point to go, because of the system and how well it's developed by all the Stack Overflow guys, I really like it. So, hopefully more people will keep coming there. You know, I'm not saying to desert your favorite board. I think All About Circuits is another really strong one. AllAboutCircuits.com is a really, really big board. But, again, you know, there's always super users and there's cultures among boards and stuff like that. And sometimes it's okay to ask one thing versus another thing.

Dave Jones: Yeah, and, you know, you'd be amazed at how many obscure forums there are out there, too, because I've done searches on Google for obscure things like land patterns, defining your own land patterns in Eagle, and wondering about what the rules are, because there's all these rules in defining optimum land patterns. And you get on some weird CAD EDA board that's all professional engineers, and it's amazing the stuff that's on there. But it's a very insulated environment because that information, if it weren't for Google, I would never even know that this board existed.

Chris Gammell: Exactly.

Dave Jones: Yeah, why would you ever go and join that?

Chris Gammell: Because it's like you always need some level of involvement in order to keep up on it, right? And you don't want your email flooded with, you know, oh, well, Mike from, you know, Sweden needed some help with, you know, X project that I don't even care about, you know. So, but, you know, you could have that, too, and you could set up tags and stuff in the Stack Exchange. So hopefully more people will be going there in the future and keep on building it up. What else was on the list for today? Let's see, I've lost the list already. Well, we were going to talk about some of the sourcing issues that are out there right now.

Dave Jones: Yeah, yeah. I just wanted to go on a little bit of a rant about it. It's okay, go for it, man. Well, I guess... Let's get a background first, I guess. Yeah, yeah. I don't know if you want to talk about it. You say you've mentioned it on the show before, so...

Chris Gammell: Well, just sourcing issues in general. I mean, like, you know, for people that are starting out with electronics or, you know, or even have been in electronics and maybe aren't in industrial, not even industrial, but like commercial situations, sometimes you don't really realize that getting five parts is okay from DigiKey or Mauser or whoever, but getting 5,000 parts is often very much more difficult. But it sounds like you have been having problems getting the five parts, even.

Dave Jones: Yeah, absolutely. And I was actually... I was talking about this online this week because I've been having a real problem getting a hold of Atmel AVR microcontrollers. And it's not just something that's been going on this week, but it's something that's been affecting me basically for all of 2010. Yeah. And the issue is just that some of the very popular hobbyist micros in the AVR line are not available. And it's not just DigiKey, but it's everywhere. If you go on FindChips or if you go on Octopart, you turn up all zeros. And this is for all the major parts like the Atmega 328s are usually out of stock, although they get little bits here and there. But just this week I was looking at Atmega USB chips, the 8U2, the 16, and the 32U2, and the 32U4. And the delivery dates on these guys are out into next year in 2011 for pretty much all of them. And I think that the only reason I really bring it up, and I think the reason why this is important, is because it is just the five chips. If you're trying to make a prototype using an AVR microcontroller, you have to be very careful about which variant you pick, because many of them showed lead times in the tens of weeks. It's not just next month, but in a lot of cases, there won't be micros available for six months. And so for someone like me who's... Yeah, go ahead.

Chris Gammell: Oh, yeah. I was thinking a couple months ago even, though, they were saying into 2011 for certain things. Yeah, no, absolutely.

Dave Jones: There are some parts that have been very hard to get a hold of pretty much all year. And so what I think a lot of people are doing, and I'm guilty of this as well, is stockpiling parts, which makes it even worse, because then whenever the parts do come in, they get gobbled up by all these people who may or may not actually need them for a project that they have today, but they're worried that they're not going to be able to get them tomorrow. And so it's a real problem for people like me who... I do freelance electronics design, and I do kits, and I do a little bit of everything. And how am I supposed to keep designing circuits around microcontrollers that I can't get a hold of? Yeah. And in my case, it makes me start thinking about maybe I should look at MSP430, maybe I should look at PIC, maybe I should look at something that I can actually get parts for. But I'm so committed to the AVR platform that it's really painful to think about switching. I'm all geared up for AVRs. I like AVRs. And now I'm getting punished, basically. And I'm not really sure what the cause of it is. I've heard a lot of rumors about fab capacities and Atmel having issues. But I appreciate knowing what the real story is. So if anybody knows what's going on at Atmel or knows somebody at Atmel, throw a little bit of information our way. We'd love to hear about it. Right.

Chris Gammell: Yeah, and it's not just micros either. Just today I was looking at this EE Times article about FPGAs, which have been having a really hard time lately too. I've had that personally. Actually, it's kind of funny because the title of the article is Analysts Fret Over FPGA Lead Times. And it's for completely the wrong reason, though. Yeah, yeah. They're fretting because the lead time is going down. So they're all worried. They're all worried the man's taking a dive. And it's like, screw you guys.

Dave Jones: I want some parts, you know? Like, screw you. Yeah, yeah. We need to write a column, engineers concerned about it. Yeah, engineers concerned about stupid analysts who want to make more money. We need some parts, dum-dums. Yeah, and I mean, I can't. There are a few things that are more frustrating than to have a circuit that you love and that you've put a lot of energy into. And then to have your supplier tell you that you can't build it because you can't get the parts.

Chris Gammell: Yep.

Dave Jones: It's like the carrot dangling in front of you, but you can't actually grab it. Right. That's how I feel sometimes. And you're right. It's actually a lot of things. And I feel like the availability for a lot of components has actually gotten better from early in this year whenever it was hard to get capacitors and LEDs and things like that.

Chris Gammell: Oh, yeah, yeah.

Dave Jones: So it's really interesting, but I'm happy that I don't have to deal with that many different product lines because it's one of these things where the probability of being able to make a complete complicated PCB where you have maybe 100 different components, if any one of those components goes unavailable, you're hosed. And I had that experience this year on a contract I was on where I needed to build more boards, and I couldn't. And it's because out of 200 components, there was one component that was unavailable, and so I couldn't build the entire board. And in this case, it was an RS-45 converter, and I was able to find a substitute. But, I mean, it's things like that that drive you absolutely crazy.

Chris Gammell: Yeah, you know, I'm always pushing vendors. First thing I always say to them when they're developing a new part is, is it a standard pinout? And, you know, they always kind of hem and haw, like, oh, well, we have our new package technology. It's like, I don't care. I want to be able to rip your part out at any time if it pisses me off. So if you're a vendor, you know, does not... And the problem is, the real problem is, I think, for, like, the analog stuff that I'm using on a regular basis is it's getting so small that it's almost impossible to standardize, and they've started patenting their stupid little package technologies. I know there's a whole science behind it, but I don't care. I want parts with standard pinouts. So someone step up to the plate. I don't care what it is, if it's LT or ADI, and say, here's the standard. Maybe even IEEE should start jumping into this stuff, you know, because it's really... it bugs the crap out of me. And I would buy more parts if they were... I would definitely target parts that are standard pinout and then go for the specs after that.

Dave Jones: Yeah, because you want to be able to second source, and maybe these days you want to be able to third source things, right? Yeah. You want to have a backup plan. And I think this happens to anybody. It happens to big companies, too. You know, they get hooked on these particular parts, maybe because a particular supplier offers something unique, and then you're exactly right. It comes in some weird, obscure package. And then whenever the parts are not available, you're screwed.

Chris Gammell: Yep. Yeah, and, you know, it's gotten to the point where the design I'm working on most recently, I've been on DigiKey and Mouser and, you know, Octopart and all those, and I just sort descendingly by stock. And even though it's not reliable, because someone could come in and buy, you know, all of their stock out, because, you know, companies are doing that, that's still the best that I have in terms of a prediction mechanism. So, you know, the number one, and the worst part about that, since you got us on this rant, Jeff, the worst part about that and what I've thought of is, what if no one's buying these because they're crappy parts? Like, oh, my God, what if that's actually the reason? I mean, I don't think it always is the reason. I think sometimes it's just certain suppliers are better than others. But, you know, that has crossed my mind that I've been buying crappy parts because they're available.

Dave Jones: Well, yeah, I know what you're saying. And I usually have a pretty good nose. And I spend a lot of time on DigiKey's site. And I do the, I filter by in stock. And I usually filter by Rojas also because I want lead free. And that usually cuts the available parts down by a factor of 10 just by checking those two boxes. But it's interesting because you can usually tell if a part, like if they have 35 units left, that makes me nervous. If they've got like 80,000, then usually if it's 80,000 and it's a low-cost part, I usually feel pretty good about it. But there are certain combinations of factors that kind of raise my suspicion that maybe it wouldn't be a very good idea to design a board around this particular part.

Chris Gammell: Yeah. Yeah, it's getting, I mean, I hope it keeps getting better. But until then, you know, like it punishes those with bad fab capacity. It punishes those with outsourced capacity. So people that are doing the TSMC, you know, the fabless design. So that's on the micro side. So people out there listening that don't know. Some companies are actually, they design in-house. So they might actually, so say you're designing a new FPGA, like an Altera or Xilinx. I think Xilinx has their own fab. I'm not sure if Altera does. But say they design this FPGA. You know, they have all their simulations on their computers and all these gate counts and they design all this fancy stuff based on the software. And then they just ship it over to TSMC, which is Taiwan Semiconductor Manufacturing Company. And, you know, that is the biggest fab company in the world, followed by Intel, I believe, and then followed by Samsung. Or, you know, all three of those are the big players. But basically, everybody puts their stuff through these chip boundaries. And now you also deal with, you know, Dave and I have talked on here before about how the Apples of the world get all the priority on parts. Well, not only that, now the people that are making the parts are also giving priority to the people that are selling the most parts. So if National Semiconductor is selling all their parts to Apple, so Apple gets, you know, 100,000 parts from National Semiconductor, and then National Semiconductor is in charge, or is in front of maybe ST Mike, well, no, ST has their own capabilities, I think, but is in charge of, is in front of some other, you know, company that's doing very similar things. And now the people way at the bottom, you know, we are back at the line, basically. And we are getting no parts. And, you know, it's like, luckily DigiKey and Mouser and all those guys are, you know, big buyers because they buy them in bulk, and then they sell them to us for a 50% markup. But it's...

Dave Jones: That's right. Yeah, thank God for them.

Chris Gammell: Yeah, I am so appreciative. Thank you. Thank you, DigiKey. We love you. Oh, so awesome.

Dave Jones: Yeah, yeah.

Chris Gammell: Yeah, that's great. Looking on, like, a vendor site, they're like, oh, 1,000 piece price is like $2. Then you go to DigiKey and it's like, oh, no, they were saying $7. That's what they really meant. That's right. That's right. Screw you guys.

Dave Jones: Yeah. No, it is a real problem. And I think that as hobbyists, people like me that actually are not just hobbyists, but small businesses, that's the real... Those are the people that get really hurt by this because we don't have the volumes and we don't have the units to be able to get the quantities that we need. We're not even on the radar for these companies. And I think that's the thing that's affecting a lot of the, especially dip packages of anything these days because I think the only people that use dips are the small companies and hobbyists for most parts. And so we get basically whatever is left over.

Chris Gammell: Yeah. And there's a bigger market for that now, too. I mean, a lot of people have... I think SparkFun has a bunch of breakout boards and Adafruit has a bunch of breakout boards and I've seen some elsewhere, too. And, you know, I think there's a market just for that. I mean, I've seen industrial socket designers, too, that do the same kind of thing. So if you're out there and you're thinking about, well, what kind of company should I start? That could be one of them, honestly. That is a big need because it's a silly little thing and you'd have to buy, you know, 10,000... You know, you'd buy, you know, a couple panels worth of PCBs and get 10,000 pieces out of it. But, you know, eventually you make your money back because it's necessary and it's going to keep happening. You know, the dip is nice to have because it's pluggable and it's going away, like you said. So we'll see what happens in the near future. I mean, I don't know if it's near future, either. I know that... I think Atmel said they were going to keep it for a while, right?

Dave Jones: I think so. I mean, I don't know. I hope that they'll tell us well before they discontinue it. But, yeah, I think it's probably something that they're not in... They'd probably like to get rid of it, but maybe there's enough people asking for it that they don't.

Chris Gammell: Yeah. Well, there's always those legacy issues, too. I mean, there's products out there that, you know, the military buys, you know, components and they buy products and they want the same components in them for 20 years at a time.

Dave Jones: So sometimes it happens. I will say that most of the more interesting parts are not available in DIP. You tend to be fairly restricted for the features because even in some DIP packages, you'll get less features than you do with the same part in a surface mount package just because the DIPs don't have as many leads. But, you know, they're easy to use. They're easy to test. And for people that just want to throw something together like a prototype, it's hard to beat the DIP.

Chris Gammell: Yeah. So what about the lead-free you were talking about earlier? You said you do your boards in lead-free. Is that a design requirement for you or is that a personal choice?

Dave Jones: Well, it's actually, for me, it's both. For boards that I've done for contract work, I've been actually required to deliver lead-free boards by my clients. And usually it's because the board is not just for the United States market. And so if somebody wants to be able to take something that I've worked on and then be able to sell it overseas, there's requirements for lead-free assembly. But I also feel that if possible, it should be a goal to try to make things lead-free. I think that these days it's fairly easy to get pretty much all the components lead-free. Like if you actually look at DigiKey, by choosing lead-free, you usually don't eliminate very many choices unless it's an old obsolete component. Most components coming from the manufacturers are lead-free anyway because there's a lot of pressure on them to be lead-free so that they can sell their parts in world markets. So really, I mean, how much effort is it to go lead-free for your final assembly? Well, you have to use lead-free solder.

Chris Gammell: See, now that's where I'm going to stop you. I knew you were going to get to that point. I'm going to stop you right there. I hate lead-free solder.

Dave Jones: Why do you hate lead-free solder?

Chris Gammell: It takes the higher temperature. It doesn't flow as nice. I mean, like, I understand lead is a nasty thing, but I love my lead-ed solder. I don't know if I sound like an old codger right now, but like, it's just a superior product, I think.

Dave Jones: Well, it certainly is, but not if you think of the big picture. I mean, it is. I agree with that. Yeah. Yeah. Yeah. I mean, it's easier to work with. It wets better. But I've actually found it, and I was one of the people that resisted using lead-free, but I actually had someone kind of force my hand into just giving it a try. And basically, I was working somewhere where there was absolutely no leaded solder anywhere in the lab. And so the only question, the only option was to use lead-free solder. And I actually found that the lead-free solder that you can buy now, and I can't remember which specific alloy it is, but it was, you know, your ordinary lead-free solder on a spool. It's like some silver tin, is it 70-30 kind of thing? Oh, no, it was more, I think it was, what was it, tin, silver, copper, and I'm probably completely wrong about that. But it was a... We do electronics, not materials, blah, blah, blah. No, no, it's actually an alloy of three metals, and it's one of the most common, and I wish I could remember it, but it's been a while since I looked at the label. But it's the most common kind if you buy solder from any of the major distributors and you pick lead-free. But the thing about it is that once you get used to it, which means having a soldering iron that has an adjustable temperature, and if you're doing like a hot plate, you need to be able to turn the board up a little bit hotter to get it to reflow. You know, your reflow toaster oven has to be a little bit higher. But once you make those adjustments, it's really not that hard to work with, and I've found that if there's any kind of a problem, usually a little bit of extra flux, for the prototyping kind of stuff, you can get away with that. You just add some flux, and then it usually flows, and it looks just as good as leaded solder. So I have to say that coming in as a skeptic that it was going to be a nightmare, I actually am not scared of it at all. And I can use it interchangeably without really having much of a problem. I still use leaded solder for some of my personal projects, but pretty much anything that I work on for a client, or anything like my kits, I assemble entirely lead-free, just because it's not that hard to go that extra mile. And then it makes me feel a little better that at least if I'm shipping kits internationally, I'm not sending them lead-containing parts, which you're not supposed to do. And I'm not actually completely versed on the requirements, but it's my understanding that if you're making stuff and you're sending it to Europe, that you've got to be Rojas.

Chris Gammell: Right. And I think Rojas is getting more stringent now as well. I know I saw something about that the other day. It was a couple weeks ago, actually. And they're adding a couple more chemicals to the list. I don't think it was... I think they've got the solder thing down pretty well. And for all of our listeners out there who have given me crap before, solder.

Dave Jones: Solder?

Chris Gammell: Solder. Solder. Solder. I catch that one pretty regularly from a friend in New Zealand.

Dave Jones: Yeah, us Americans like to say the solder. Solder, you know?

Chris Gammell: Solder, yeah. Yesterday, I was going out to the supermarket and picked up some solder.

Dave Jones: Get some solder and have a yard.

Chris Gammell: Yeah, well, I don't go there.

Dave Jones: Yeah, yeah. I know what you mean. And actually, I just saw something not that long ago. And I can't remember where I saw it. But they had just added another chemical to the prohibited chemicals list. And I can't remember what it was. But it's funny because they slashed these chemicals off the list. And then you have to think about, well, what are all the things that use these chemicals that now are going to have problems? And I think that whenever they first introduced a lot of this stuff, it was really hard for manufacturers to comply. Because electronics have always contained all sorts of interesting, you know, brominated epoxies and things. Fire retardants. That's a nice way to say it is interesting. Well, not the kind of stuff you'd like to put in your body, I guess.

Chris Gammell: Yeah. I'm trying to find the chemicals right now. I know there was one on there that I definitely used before. It wasn't sulfur hexafluoride, was it?

Dave Jones: I don't think so. SF6? I think it was something more basic, wasn't it? Like cadmium or something.

Chris Gammell: Oh, cadmium's nasty. That stuff will poison the crap out of you. Absolutely. I think SF6 actually is on the list. Have you ever seen that stuff before? No, I have not. That'll be a great thing to link in. You know, we always have links on the page. But SF6 is a gas. It's an inert gas. And I used to use it. I used to work in a fab. And I worked in dry etching. And so people who don't know what that is, it's basically when you're making a silicon wafer and, you know, you're trying to etch patterns onto it. First you lay it on photoresist and then you use light to expose it. And then you use chemical etching. You either use chemical etching, which is called wet etching, or dry etching, which is like you create a plasma with like RF energy and all this crazy stuff. Well, that's one of the chemicals we used to use. It is pretty common throughout the industry. But if you look at videos online, and it is pretty nasty once you actually like hit it with some RF energy because it breaks off all the fluoride ions and free radicals and it starts to eat at the actual wafer. But if you look at it, if you look at the gas in a natural state, it's actually heavier than air. So there's a video on YouTube. There's a couple of videos, I think, where there's two things you can do with it, neither of which I would suggest because I still wouldn't play with this stuff if I had it. One is if you put SF6 into a tank, like a fish tank, you can actually, they have a video and there's a tinfoil boat that's floating on top of it and it looks like it's floating on nothing because... I have seen this. I have seen this video. It's very dense gas and that's why it happens. So it sits in the bottom of this container and it looks like it's floating on nothing. Very cool. The other thing that you can do with it is much like helium raises the pitch of your voice, SF6 lowers the pitch of your voice and... But you wouldn't breathe this stuff. Oh yeah, they do.

Dave Jones: Oh wow.

Chris Gammell: It's inert in a normal state. I mean, there will be some natural radicals from like gamma radiation or whatever's coming through the atmosphere but like I said, not the healthiest thing to do. Just like helium, you know, like whenever you suck in helium and make a funny high-pitched voice, not a good thing either. But seeing someone else do it on YouTube, you know, that's always acceptable. So I'll be sure to link that in too. And I think that's on that Rojas list. I think that's on some list and, you know, there's a lot of chemicals out there that are pretty nasty chemicals. So it's... That's one of them. I mean, like it can be very nasty.

Dave Jones: Yeah, that's interesting. I didn't know that you worked in a fab. That sounds like a pretty cool job.

Chris Gammell: It was all right. It was not as electronics-oriented as I wanted it to be. So that's why I kind of got out of it.

Dave Jones: Yeah. I actually had a chance to visit a fab whenever I was working in California. And it was actually the same fab that's in the Postal Service Such Great Heights video. Interesting. There's a music video. Yeah. And it's by the Postal Service. And it actually... The whole video was filmed inside of a fab in Newbury Park, California, that I had the privilege of getting to see the inside of. And my biggest beef with it was that it took me like 20 minutes to get suited up to go inside. And it was so hot inside the suit that I had to wear that I was constantly fogging up my glasses. Oh, yeah. So maybe I'm not cut out for working the fab. But it was fascinating to see everything that's in there. And they were showing me ampoules of gold that they used to do sputtering and all sorts of stuff. It was really, really cool. This was a gallium arsenide fab. So it's a little bit different than what you used to work with. But really, really cool stuff. And the number of amazingly dangerous chemicals that they use is really mind-blowing.

Chris Gammell: Oh, yeah. Yeah, definitely. I recall not so fondly seeing pictures of what hydrofluoric acid would do to people.

Dave Jones: Yeah, yeah. That was part of the training. In university, they told me not to spill that because I guess it goes through your tissue and eats your bones on the inside.

Chris Gammell: It actually eats the tissue from the inside. So you'll spill it on yourself. You won't feel it. And then your hand will start to rot off while you're looking at it. Wow. Yeah, it's like a chemical burn that comes from the inside. So if you even think that you've had it splashed on you, they'll put this copper paste and it leaches it out somehow.

Dave Jones: You probably made three listeners that work in fabs just considered leaving their risky occupation.

Chris Gammell: Oh, I'm sure. If they work in a fab, they know about it. I guarantee you that. Okay, good. Good. Yeah.

Dave Jones: It's good.

Chris Gammell: You know, it's really cool in there. And, you know, I like the people I work with a lot. And it was a really cool experience. But like I said, fab life is a different life. And if you want to do it, you know, if there's people out there considering it, definitely try it out. But, you know, I was working in a manufacturing capability. I was not doing any research in there, too. So that's another, like, I'm sure that gallium arsenide fab you went into was very research oriented.

Dave Jones: No, it's a production fab. Oh, okay. Well. It is. They use, gallium arsenide is used in most high power RF chips. And so this is a fab that actually cranks out parts that are in a lot of cell phones around the world. And so it was kind of cool to just be amidst. And that's actually one of the things. You were talking about TSMC earlier. There are not very many gas fabs outside of the United States. And so gas is one of those things where you still have a lot of activity, manufacturing activity going on within the U.S. And I know there are fabs outside of the U.S., but it's not like silicon where most fabs are not located in the United States. Right. So it was interesting to actually walk around in a place where they're actually making volume. And it gave me an appreciation because as an engineer, it was very easy to yell at the guys in the fab whenever they screwed something up, like patterned a wafer wrong. Or the worst one was whenever they dropped things because gallium arsenide wafers shatter. And so you end up with a million little bits and maybe you get one or two of your chips off of. But it gave me a new appreciation for how ridiculously complicated and difficult the job that they had to actually make those devices were. So you don't really appreciate that until you go there and you see that. Yeah.

Chris Gammell: You know, there was a really interesting comment. I'm not sure where it was. It was on the suggestions maybe of The Amp Hour page. And someone actually suggested that, you know, Jerry Ellsworth started doing MOSFETs in her house.

Dave Jones: Yeah, I'm familiar with this. Yeah, she was – and you should post a link. She's got some YouTube videos, I think, about it. Yeah. But, yeah, she was actually making FETs in her garage. And she was using household chemicals, like the kinds of things you'd get at Home Depot. It turns out you can buy dilute hydrofluoric acid at places like Home Depot, which is a hardware store for anybody not in the U.S. Oh, yeah. Good point. But you can go down to your local hardware store and you can buy things that contain – I think she was using stuff that contains arsenic and phosphorus. And she was basically getting dopants out of household chemicals. And she actually patterned some FETs and they worked. Yeah. So that was really, really cool to see. I actually got to see that at the Maker Faire last year. She was there. And so, yeah, there's – I think it's a far cry from having a Home chip fab because she was making just a couple of devices. But it's interesting because whenever you think about it, at the beginnings of the semiconductor revolution, people were working out of what wasn't really much more than a garage. Right, yeah. So if you – I always like to think of things, if you trace things back far enough, there was a point where a guy was doing that in his garage. Yeah. So it's funny how far away we get from that. But there was a time whenever it didn't take a lot to make fairly low-performance devices. But I think it's interesting because people like Jerry are trying to retrace those steps. Right.

Chris Gammell: Well, the thing I was going to bring up, too, is that someone had mentioned – I'm trying to find it right now. They mentioned the fact that this actually might be possible in terms of, like, photo resist etching and stuff like that. Like, I think he said – I hope I find it, too. I'll at least post the link. But, you know, basically, you know, modern laser printers have decent resolution down into the micron. You know, like 20 micron, I think he said, was the lowest resolution you can get right now. Wow. And, you know, that's a long – that's a far cry from where they are now. You know, 20 microns is 20,000 nanometers. So they're rapidly approaching 1,000 times smaller than that. I think the smallest geometry they're producing right now is in 30 nanometer for, like, the Intels and the Samsungs of the world are doing, like, very small flash memory and microprocessors – or, sorry, CPUs or whatever. But basically, you know, if you're starting to start moving up in the geometries, this is actually a possible thing to do. And it's a really interesting notion. I never thought of that before. But, you know, much like we can do the – you know, like the ink hack where you can print out a piece of paper right now and then kind of push that onto a piece of copper and then use that as your etch resist if you're etching a board at home. Have you ever done that at home?

Dave Jones: No, I haven't. I've always kind of avoided it just because of the chemicals.

Chris Gammell: Not worth it. Right, yeah. I agree. But I've seen that before where people can – you know, you can actually print out – you can print out your pattern on just a regular paper. Then, you know, you kind of transfer it over to a copper-clad board with, like, FR4 below it. And then you just etch the copper. And to be honest, folks, that's basically how you make a semiconductor just on a much, much smaller scale. Yeah, yeah. And you layer it too. You know, you've got to layer it and grow stuff and dope stuff. So obviously there are a lot more steps, but, you know, like –

Dave Jones: Yeah, you need a furnace. That's maybe one of the hardest things about it. You need to be able to get things really hot. I got an oven. No, I mean really hot. I got two ovens.

Chris Gammell: Well, yeah. I don't have two ovens. But, yeah, you know what I mean. Like, there's – I think that eventually there could be – you know, much like the – there's laser printers or laser cutters now that are open source. And there's, you know, the 3D printers at open source. I think it is possible that eventually very stupid chips could be made at home. You know, Jerry started on that path. But I think that, you know, in terms of a –

Dave Jones: Yeah, and I've actually thought about this a bit over the past few months. And one of the things that – and here I go – I'm going to be talking about these hacker spaces and things like that. So I guess whenever I'm on the show, it kind of – the topics change a little bit to more of the open source hardware, you do-it-yourself kind of stuff. But I think it's interesting to start thinking about what would happen if do-it-yourself people, hobbyists, people in hacker spaces started thinking about making their own chips. And what I'm actually talking about is not making them at home in your garage because, to me, the cost in equipment and chemicals and all that and the hazards of working with these things are kind of a problem. But there are services which traditionally universities have worked with to make chips. And Moses is the most famous one. And that's where a lot of students that are in undergraduate or maybe grad school – I never got a chance to do it whenever I was in university. But you get to make a chip. And it's like a three- or six-month-long process of designing the chip, laying it out. And interestingly enough, a lot of the chip layout tools are open source. Really? Exactly, because all you're really doing is putting down squares to define gates and things like that. And you can get some sort of okay simulation tools now. And so it would be interesting to think about actually having a group of hobbyists or maybe small businesses, people that are entrepreneurs, basically pull together to get a mask actually fabricated. And Moses, I think, is in some fairly decent processes. It's not like the latest stuff. But you don't really need the latest and greatest things for most designs that you'd want to do, especially if you're working on analog circuits. Those tiny gates are not helpful for analog. They actually hurt you more than they help you. Right, the leakage. Yeah, and also voltage range and things like that. Yep. The breakdown voltage is really, really low for those advanced processes. And here, we still like working at 3.3 and 5 volts, right?

Chris Gammell: I like higher than that, I'll say. I love plus minus. If those vendors out there that are making me standard, if they also want to give me plus minus 15 rails, I will take them. Just a side note there.

Dave Jones: You like the op amps from the 70s and things like that, right? Let's say the early 90s. Let's say early 90s. Oh, okay. Okay, well, so I think it would be interesting just to see what would happen if you enabled people to make their own chips. And even though it's not at home, it's kind of the next best thing. Think of it like batch PCB, but for integrated circuits. Yeah. And the costs are quite high. It ends up being, I don't know, it's in the tens of thousands of dollars. So it's not like the affordability is there. But I just, I start thinking about, well, so if you had this capability, what would you do? And I guess I'm not entirely sure. But it would be interesting to make some very highly integrated circuits that you couldn't do using traditional ICs. Like there's a lot of RF circuits you could make that if you had access to your own IC process, you could integrate an entire radio on an IC. Of course, yeah. So there's things like that that I think would be interesting to play with. So I don't know how practical any of this is, but this is something I daydream about. And it'd be interesting to see what happens down the road as, you know, there was probably a time whenever printed circuit board technology was considered totally out of reach for the hobbyist. You know, where something like batch PCB would never have existed. So, you know, what if down the road we have, I don't know what you'd call it, batch silicon or something where people can actually do their own designs. You can call it batch C. Batch C. That doesn't sound right, no. No, no, maybe not. We'd have to think of a catchy name, otherwise it'll never catch on. Yeah, of course. Who else would do it otherwise, right? Yeah.

Chris Gammell: Yeah, I don't know. It's an interesting thought. I mean, the, yeah, the variation between parts. I've daydreamed about it too only because, you know, some of my job is to replace obsolete components. And when there's not an actual component to replace it, you know, if you actually look at parts these days and say you have a circuit that does some function, you know, it turns on bit A and that turns off bit B. You know, something actually a lot more complicated than that. But if you go and try and design that in, you know, JFETs or even BJTs or something like that, all of, you know, if you look at a BJT package right now, coming in like a standard package, you're not going to be able to even come close. Because if you look inside the package, then it's actually the leads are then wire bonded to the actual chip. And the actual chip, it looks like a speck of pepper. So, you know, you'd never be able to match the actual size of what you, what you, what Jeff, what you and I use for discrete ICs or discrete transistors, you would never be able to use those to build up a, you know, an op amp in any usable manner at all.

Dave Jones: Yeah, well, I mean, it depends on what your space requirements are. Some of the early op amps were actually made of discrete components, which blows my mind. They were modules, right? Yeah, that completely blows my mind. And I guess it's Phil Brick and those guys way back, 40 years ago, I guess, more than that?

Chris Gammell: Yeah, 69, I think.

Dave Jones: Yeah, so they were playing with op amps made out of discrete components. But I agree, to do anything practical and have it be small and anywhere near the footprint of the part you were replacing, I think that's where FPGAs and things like that come in. Right. Because you can do a lot within FPGA. But if you're trying to do analog circuits, then that doesn't work as well. And that's what led me to start thinking about this Moses stuff is that there's not really any options for doing integrated analog. There aren't very many analog FPGAs yet. Right. Right.

Chris Gammell: Well, yeah, there's that FPAA, but that thing is not what I like. Interesting. I think it'll be interesting because the, you know, one of the things that really limits people right now, because I think the batch PCBs, what those came from is people going out and buying what used to be high-end equipment, you know, keeping it alive, having a good, you know, technician around that can keep it alive, and then pushing technology onto that after the fact, right? Yeah. So they might go out and, you know, buy board processing equipment after the fact. And there is definitely investment there, but it's not as much of an investment as it was when that industrial equipment came out.

Dave Jones: Mm-hmm.

Chris Gammell: The thing about a chip fab, you know, you need so much of it, and I'm not sure how much it actually sticks around because a lot of the internals of those are actually replaceable. So, like, inside, like, a dry-etch chamber, it's actually all replaceable parts.

Dave Jones: Yeah. Yeah. Well, I mean, Moses is actually, those wafers are run at TSMC for if you're using a TSMC process. Oh, okay. What they do is they slip the wafers in, you know, at probably 3 in the morning or whatever the equivalent in fab time is. It's probably during their slow season or slow runs. Yeah. They slip wafers in. So you're actually getting right on the actual equipment. So it's a little bit different than, well, actually, no, it's not different because batched PCB, as you might already know, their back end is at Gold Phoenix, which is a real PCB manufacturer. Oh, I didn't know that. So they're not actually, SparkFun doesn't actually make their own boards. And this is all public knowledge. It's been talked about a lot before. They actually send the boards out. So what they're doing is they're providing a service. So they're panelizing designs from multiple contributors, and then they're sending the panel out because it's not cost-effective to get one board made. It's only cost-effective if you get a panel. And so they actually panelize them. And that actually reminds me, I actually have a shout-out this week, which I didn't think I was going to be able to say, but now that I'm on the show again. Hey, hey. Go for it. I wanted to give a shout-out to Dorkbot Portland. Oh, yeah. So for the international guys, this isn't going to be super helpful. I can't imagine the postage is going to be great. Although you might want to look at it anyway. There's a really, really helpful guy. And I actually can't remember his real name, but his handle is Lane. And if you go to pcb.lane.org, that's L-A-E-N. We can put a link on the site. What Lane is doing is he's actually panelizing boards just like Batch PCB does, but at about half the cost for most small boards. And I actually had some boards made there a couple weeks ago, and they turned out great. So for guys that are looking to do prototypes, it's kind of a cool service. And it's much more targeted towards hobbyists and people that need prototypes for doing stuff at home than Batch PCB is, because it's super, super low budget. But you pay a lot less to get more boards, basically, is what it comes down to. So it's kind of a cool service. And the reason I'm kind of advertising for it is that one of the things that keeps it going is getting enough contributions to fill PCB panels. And so if there aren't enough people that want to do boards, then the service stops, and then people like me are unhappy.

Chris Gammell: Right. And he actually was the one, I don't know if you saw that Adafruit post a couple of days, I think you did. It was just the other day about one of the silkscreens that was on that most recent order was like some beautiful silkscreen program.

Dave Jones: Yeah, yeah, yeah. Yeah, so because he gets such a diverse set of contributors to the service, he gets some really crazy stuff. And so you end up with boards that have these really ornate silkscreens. And the downside is that he's told me that sometimes what will happen is he sends the board in for manufacturing, and the PCB manufacturer will call him back and say, what is this? You know, are you sure this is what you wanted to do? Because, you know, obviously PCB shops are not used to seeing artwork at the level that these boards are on their PCBs. So it's kind of an interesting thing. But, yeah, I've seen some really cool boards come out of his service before. So, yeah, it's really cool. Yeah, that's awesome. That's really awesome.

Chris Gammell: And we'll post a link, too, so you can get to that. And, you know, thanks to, I think it's James, actually. James, yeah.

Dave Jones: Yeah, that's right. That's right.

Chris Gammell: So thanks to James. I don't know if he goes by James, but that's his given name.

Dave Jones: Yeah, so hopefully I'm not in trouble for advertising his service on the show.

Chris Gammell: Oh, you know, our sponsors will be so pissed. Oh, we don't have any. Oh, I see. Yeah, don't worry about that, buddy.

Dave Jones: I mean, I wouldn't do it if I didn't consider this something that's of value to people that are doing electronic design. That's right. Well, you know, the good stuff's got to rise to the top.

Chris Gammell: So that's a good example of it. That's good. I'm really glad that he does that. I haven't done it myself. I'll have to try that out.

Dave Jones: Yeah, and I don't. He hardly even makes money on the service. I mean, he's doing this just to help people get boards. So it's kind of something he does out of, I don't know, generosity for the community. So it's kind of cool. And, I mean, the benefit is that if more people start using the service, he can run more panels because right now he's running one a month. And so sometimes you've got to wait a little while, but for the price, it's really, it's totally worth it. So have you done it? Yeah, I have. And the boards are beautiful. Yeah. I mean, really, really actually better quality than I've seen from Batch PCB. So the boards I got were really cool. One of the funny things is that you don't necessarily know what solder mask color you're going to get because I guess the guys that are doing the boards just throw these in. Just like I was saying, they put them in whenever they have capacity. And so you get what you get. And these came back purple. Awesome. Which I've never seen before in a PCB. But it's kind of neat. And I was actually very happy. So I wouldn't advertise the service if I hadn't had a good experience with it. But I was very, very happy with the boards. I've been playing with some USB AVRs. And so I needed boards. And now I've got them. And I got them within two weeks. So that was pretty good. Yeah. You mentioned that last time. So that's going pretty well. Yeah, it is. Although, like I was saying, the AVR availability is a real killer. And that's definitely impacting me on this design. And it makes it very hard to think about designing more things with AVRs. So that's why I wanted to talk about that because it's something that affects me on a very real level is I just can't get parts. Yeah.

Chris Gammell: That's too bad. Actually, I found the name of the person that was mentioning making the IC at home. So it was Richard Neenhouse. Neen Kuss. I'll post his name as well. I think we might have talked about it at home. But he was mentioning that the Intel 4004 was built on 10 micron technology. The 6502 was built on 3 to 5 micron technology. Micron being a micrometer, if you don't know that out there. And he said, current laser printers are hitting 1,200 DPI, which corresponds to a bit over 20 microns. So maybe give it a couple more years. I don't know if there's any real impetus to move to higher DPI because I know a lot of documents are still at 300 DPI.

Dave Jones: Yeah, that's right.

Chris Gammell: But maybe it's capable.

Dave Jones: Yeah, and I almost wonder, you know, definitely good luck with that project. But I have to wonder, is that 1,200 DPI really real? Oh, right. That's one of those things that I've never believed that for one minute.

Chris Gammell: Who watches the Watchmen, that kind of thing?

Dave Jones: Yeah, exactly. I mean, who's actually verifying that 1,200 DPI is 1,200 DPI? But I say if it's true, then that's pretty killer. And so I guess he's waiting for like the 2,400 DPI, you know, because then he'll be right up at the resolutions of some of those old chips. Yeah. But, yeah, if somebody manages to make a 6502 in their garage, my hat's off to them. That would be pretty cool stuff.

Chris Gammell: That would be cool. All right, well, I think that's it for us. I know you said you had to run soon, so I think...

Dave Jones: Yeah, I've got to run, but it was great to be on the show again. Yeah, thanks for the last minute. Jump on on here. Yeah, thanks for having me back, and I'm just happy to talk about electronics and happy to be on the show.

Chris Gammell: And if people haven't looked at Jeff's site before, it's MightyOm.com. His Twitter is Twitter.com slash MightyOm. That's how Jeff and I keep in touch. And like he said... Oh, goodness. I've already forgotten his name again. We talked to James, too, on there, who goes by Lane. But it's actually kind of cool. And like we've kind of brought up a couple times on this show before about the critical mass, and we mentioned that with Chip Hacker today, too. The more people that participate, the richer the experience it is for everyone. Yeah, absolutely. Don't be afraid to jump in, and we'll hope to talk to you on there. Excellent. All right, cool. Well, Jeff, I'll talk to you soon, and everybody else will hopefully talk to you next week. All right, sounds good. All right, bye. Bye. Bye.