Pick, Place, Podcast
Understanding Controlled Impedance w/ NCAB Group Field Application Engineer, Ryan Miller
We apologize for our extended break! Chris has been hard at work from morning til late at night every day setting up Worthington's two new Fuji pick and place machines. Now that some of the dust has settled we hope to get back on a more regular recording schedule.
We've had several requests from our listeners to have an episode on on impedance control, which is unfortunately a topic Chris and Melissa have little expertise on. Fortunately we were able to connect with Ryan Miller from NCAB group who specializes in the subject and kindly agreed to come on the show to give us a lesson on what impedance control is, in what situations is it necessary, and more.
pickplacepodcast.com
We apologize for our extended break! Chris has been hard at work from morning til late at night every day setting up Worthington's two new Fuji pick and place machines. Now that some of the dust has settled we hope to get back on a more regular recording schedule.
We've had several requests from our listeners to have an episode on on impedance control, which is unfortunately a topic Chris and Melissa have little expertise on. Fortunately we were able to connect with Ryan Miller from NCAB group who specializes in the subject and kindly agreed to come on the show to give us a lesson on what impedance control is, in what situations is it necessary, and more.
pickplacepodcast.com
Gotta shake off the cobwebs here,
Melissa:Right,
Chris:Shh. All right. welcome to the Pick Place podcast, a show where we talk about electronics, manufacturing, and everything related to getting the circuit board into the world. This is Chris Denny with Worthington.
Melissa:this is Melissa Hough with CircuitHub.
Chris:Welcome back, Melissa.
Melissa:Welcome back Chris. It's been
Chris:been a long
Melissa:a long time.
Chris:I, I, I feel bad for I feel bad for taking such a long break, but let's be honest
Melissa:It was important
Chris:nobody's gonna miss this show. This is, Yeah. I mean, my mom
Melissa:your grandma,
Chris:you know, my grandmother, they're gonna be, they're gonna be disappointed they didn't hear from me. But otherwise, I think, I think nobody's gonna miss this show.
Melissa:but you were doing very important things as I'm sure most of, probably all of our regular listeners know Worthington recently took delivery on two Fuji pick and place machines. So Chris has been busy, busy, busy.
Chris:Yes. Yes. Here has been my schedule for the past, like month and a half. Wake up at six. Take the dogs out and work on my laptop until I don't know, eight o'clock or so. Head into the factory, work there till six, seven o'clock. Come home, have dinner, see the wife for 30 minutes. Then hop on my laptop again till midnight. Rinse and repeat. just over and over and over and over and over for like the past six weeks. It's just been, it's just been crazy. It's very, it's been very exciting though. We we have for a long time struggled with we knew we had to get new equipment and, and we took a long time to do that research for that new equipment because we knew we wanted to set ourselves up for the next decade. With the right supplier and the right technology, you don't make a decision, a 10 year decision like that lightly. So, and we actually have had whole episodes talking about why that's important, but in the meantime, your old equipment is suffering and you don't wanna invest in it. And so you're just limping by So it's been, that's been quite a transition. And we're doing this transition, you know, mid flight, so to speak, because we don't have floor space to just drop it on our floor and slowly introduce it. We actually had to swap out the old equipment for the new equipment and in real time try to get up to speed with it as well. And it's it is been really challenging, but we have a fantastic team. Fuji as an organization has been phenomenal. Their support has just been second to none. I've worked with a lot of equipment manufacturers and these guys are absolutely top tier. been absolutely wonderful to work with them. So really, really, really pleased with how everything's going and really, really, really pleased with the equipment itself., I don't know how to describe just how perfect the boards come out of this machine. I guess we've gotten so accustomed to seeing boards a certain way, and then when you make this switch all of a sudden, like things are placed perfectly. I mean, per like, freakishly perfectly. I can't understand how it can be done so perfectly. And yet they are, and, and each and everyone, and we're, we're doing batch after batch after batch. And I'm working with the quality control team and I'm asking them like, Hey, what are you seeing? Are you seeing any issues? Like, is there any particular thing we should be aware of? And they're like, No, nothing. Nothing. There's, there's no rework. There's
Melissa:They must be happy
Chris:just, Oh, they're thrilled. They're thrilled. I keep joking that I'm gonna buy all of them some work gloves so they don't get blisters from all the sweeping and mopping we're gonna have to have 'em It's just been, it's been absolutely fantastic and a huge. A huge sense of relief because, you know, you're, you know, you spend a million dollars, you wanna make sure that it goes well. And so far it's, it's very, very promising. So yeah, we're very, very excited about it. And it's just more to come. Hopefully we can get back to a, a regular recording program now that I can, I can spend a morning here and there to sit down and, and record an episode of the podcast and get back to it because I'm sure unlike I like to joke, I'm sure some people do miss it. And we love all of you and are grateful for all of our listeners. Very much, very much grateful. How about you, Melissa? What's ? What's new with you?
Melissa:I've been not editing podcast, which has been, which has been a strange break cuz it's become something that's just such a significant part of my schedule. Um, yeah, so just been finding new projects to work on, focusing more on improving our website and documentation and such, so, which is I'm sure always needed.
Chris:Isn't that a great feeling when you're like, you're, you're able to get to like, Oh, I've wanted to make this change for so long on the
Melissa:Yeah. Like the to-do list that's been there for years sort of thing.
Chris:Exactly. That's fantastic. Melissa had a shout out to listeners a couple episodes ago, or I think we've been doing it for a couple consecutive episodes of asking for listeners to write in with show suggestions, and they did.
Melissa:Mm-hmm.
Chris:It was fantastic. We heard from Jesus, Sarah and Johann. I, I apologize if I butchered your names at all, , but they came up with some awesome, awesome show
Melissa:Yeah. Some really long lists there.
Chris:really long lists
Melissa:Yeah. cuz we needed it
Chris:we needed it, we're starting to run out of ideas and I'm like, I don't know. I don't know how much longer we can do this podcast for. It's gonna have to turn into an interview show. And then when we got these suggestions it was like, oh no, there's still a lot to talk about here. So,
Melissa:yeah, So thank you. Thank you very much.
Chris:Yes, and please continue to write in a regular request that we get from listeners has been to have an episode on controlled impedance and, and designing for that sort of thing. And rather than try to talk about something I know nothing about we were working on trying to find somebody who does know something about it. So we asked around and ended up talking to one of our primary PCB suppliers called NCA B or I don't know if they pronounce it, NCA or N NCAA b I pronounce it N NCAA B. Anyway and we got, we got an expert on the show. We wanna introduce Ryan Miller, who is a, I believe, Ryan, correct me if I'm wrong, but you are still a field applications engineer for N C B, is that correct?
Ryan:Yeah, Chris, that's correct. I'm a field applications engineer for NCAB group. It stands for Nordic Circuits Incorporated.
Chris:Nordic as in like Norwegian, like Northern,
Ryan:actually we're based in Sweden
Chris:No kidding. Oh, okay. So not Norwegians. That's Swedish.
Ryan:we, we have factories in, in Europe here in the United States and throughout China and Taiwan. So,
Chris:Yeah
Ryan:you know, we're. Pretty big company, but we don't want to be the biggest company. We just wanna be the best. We wanna make sure the PCB is delivered on time and that is accurate.
Chris:sure. Right on. And, and I believe if I listeners may or may not be aware of this there's a, there's a few different ways that circuit boards can, can be ordered by customer, by people like Melissa and myself. We can go directly to a PCB factory and work with their engineering department, these sorts of things. But a lot of times the PCB factory doesn't always wanna work directly with designers or assemblers. A lot of times they, they the, the business models have been set up to have more of a brokerage model where you have you have like a team of really highly skilled engineers who can take, take data understand it, you know, and, and find the ideal factory to produce it. And if I, if I understand correctly, that's kind of, I'm oversimplifying it, but that's kind of NCAA b's business model. Is that, is that right Ryan?
Ryan:The, Yes, you you hit the nail on the head. The advantage there is you know, we get to very carefully choose and select the factories that we work with so we can make sure we're getting the best quality board possible. And they we have a factory management team that audits factories for a year more before they make it to our primary selection list.
Chris:Okay. Wow. Yeah. Nice. And it's, it's a nice business model because it means we can, like, let's say let's say I got a two layer PCB that's gonna go into a, you know, real simple business to consumer product that's a toy or something that doesn't have to last a whole long time. I need to get the cost as low as possible on that business to consumer costs. You need them as low as possible. So, NCAA B'S team they'll find the, lowest cost. You know, it's, oh, it's a two layer, it's all 10 mil traces, no, no issues there. Versus if I have a 16 layer. Board with stacked vias and, and controlled impedance that's gonna be a different factory.
Ryan:Yeah, we can, we can go all over the place from the two layer boards that go into toys all the way up to the HDI boards with, you know, I, I like to call 'em the beast. You know, they have blind buried vias and via fill controlled impedances.
Chris:Yeah. Yeah, exactly. Well, speaking of controlled impedance, that is that is what brings Mr. Ryan to the Pick Place podcast today. This, this topic of controlled impedance. I, you know, I have a friend who he, he helps to run a factory and he knows next to nothing about controlled impedance and this. This friend of mine would really like to know more about it. And because when he is talking to his customers he doesn't know what controlled impedance is and he doesn't wanna sound like a fool. So, for my friend if you could maybe explain to our listeners, maybe at like a very generic level and then we'll get into details. What, what, how would you define what controlled impedance is? Cause my friend really, really doesn't know.
Ryan:Well, I think what, I'll tell your friend and I think it's important to start at the question of why do we need controlled impedance traces?
Chris:My friend would like to know that.
Ryan:yeah, well, there's a few different reasons. One is, is you know, we need impedance matching for antennas.
Chris:Antennas okay.
Ryan:Another reason is maybe for a differential pair for a USB connection or another reason is like maybe an FPGA or a component needs propagation delay in the signals to make accurate calculations. So there are
Chris:back to those crypto guys with their f PGAs.
Ryan:Yeah. Yeah. And, and, and I'm, I'm no expert on components, so I don't know all the components on the market that
Chris:Yeah. No worries.
Ryan:But you don't have to. That's the beauty of controlled impedance traces. I, I don't think they're, they're too difficult to understand. They're just misunderstood.
Chris:Okay.
Ryan:There's not a whole lot of training out there about.
Chris:when you say that like, Let's, so you mentioned three specific things. And maybe we can, I'm sure there's much broader applications for this than those three specific things, but you mentioned antennas USB signals and fpga something very fancy propagation anyway but in, in antenna, you, When you say in antenna, you mean like maybe literally something like a Bluetooth antenna or, or a wifi antenna, or are you talking like something more high end custom job for, for that sort of thing.
Ryan:It could be either one.
Chris:Okay.
Ryan:I've seen circuit boards with the antenna just etched right into the board and it's gonna stay at, you know, right there. Or it could be something RF or something like that, that's gonna need an external antenna connection.
Chris:Yeah. Interesting. Do you know the why's as to like, if I'm designing an antenna or I, I have this trace on my board that is gonna act as an antenna. Why does it even matter? What, why can't I just stick a piece of copper down and, and hope for the hope for the best?
Ryan:Okay, well, I'll, I'll try to explain this. From what I remember back when I was in Signals and Systems class when I was getting my degree For the, the RF signal to make it from one location, you know, from the transmitter to the receiver. The impedances have to match from the trace all the way to the antenna itself. And the antenna and the traces. They have to be the same impedance.
Chris:Okay.
Ryan:I, I believe that is to reduce reflection.
Chris:Yes, I've heard about this, this reflection concept where it, like you, you almost get this, you know, my, my little pea brain makes, it, makes me think of like water sloshing in a, in a, in a bucket or something where it like pushes back on itself when it's like, no, it really needs to go the one, the one direction, but somehow it's reflecting back.
Ryan:yeah. And, and you know, after it leaves the receiver, it's at the mercy of the environment around until it gets or after it leaves the transmitter.
Chris:Yeah.
Ryan:It's at the mercy of the environment till it gets through the receiver and on the receiver end. That antenna and impedance trace and those connections, all the impedances have to match there as well. I, I'm not gonna pretend to know anything else about radio frequency transmission. I have no idea.
Chris:Don't, don't, don't apologize for that at all. That's, that's totally fine. Yeah. I, I just, and I didn't want to get into the specifics of it, but at least at a general level because honestly, I, I've never understood what what it's used for. Right. Like when you're in manufacturing, and obviously you know this too, it's like, I'm just trying to figure out how to make it, I don't know what it does. It could be a concrete life jacket for all I know. I just , somebody gives me a print and I'm trying to build it
Ryan:Yeah, that's, that's where my job is easy. I, you know, I follow the Gerber data, the specifications that the customer sends, and if there's a problem I'll just go ask about it.
Chris:Yeah, exactly. Well, one of the nice things you guys have done for us in the past is we had a customer who had, they got a quote from you guys that was like, hundreds of dollars for this board. And I looked at the board and I'm like, This doesn't look anymore different. Than any of these other boards I've ever built in the past. I don't know why it's hundreds of dollars per board and my customer's, like, I I was expecting to get these for less than like 30, 40 bucks a board. Well, once, once you guys took a look at it and you looked at the data, you're able to say, Well, look, you've got all kinds of weird design issues here and you know, I don't know if these are deliberate or not, but this is what's driving all the cost. You've got these stacked vias and you've got these blind vias and you've got these fills and I don't know that you need all that if you can eliminate and, and that NCB and our customer worked closely together on it and drove just completely eliminated all those high cost things on the, on the design, and it was a huge help. So yeah, sometimes you just look at Gerber data, but also you look at enough of it that you start to go, Ah, I don't think you're doing this right,
Ryan:Yeah, I, I've looked at Gerber data you know, before I started working for NCAB group, I actually worked for a PCB manufacturer for 10 years. And nine of those 10 years, I Did nothing but DFM analysis on rigid boards.
Chris:Oh, cool.
Ryan:And so I've seen a few sets of Gerber data. You know, when we see something on Gerber data that looks questionable, I always try to give the, the designers a benefit of the doubt.
Chris:Yeah.
Ryan:you know, I, I've met designers from all walks of life and all skill levels
Chris:Yeah.
Ryan:I always try to do my best to help the designer reach their goal. So, if, if there's one thing that I constantly see, something that could be changed or something that even needs to be changed consistently as controlled impedance traces and stackups.
Chris:Okay. Wow. That's fa Well that's, that's, that's great to know then. So, I guess that leads me into my next question very well. When, when somebody is sitting down and they, let's say they've got how comfortable are you talking about usbc? Cuz I feel like that's something that probably requires a, a fair bit of controlled impedance on those things.
Ryan:Yeah, I can, I can talk anything controlled impedance, but my knowledge pretty much, pretty much terminates at the termination of the traces,
Chris:Sure. That's totally fine. I don't, I don't wanna necessarily talk about the, the stack up of of, of what, what's the word I'm looking for? You know, we're not the IEEE here, we're just, all I wanna talk about is, let's say I've got a USBC design
Ryan:Mm-hmm.
Chris:and I've got controlled impedance, cuz I've got differential pairs on this USB design. If, if, if I was laying out this board as. An idiot that I am, When it comes to PCB layout, I would just draw two little traces from whatever, you know, my USB chip is on the board and stick them into the USB port. And I would probably completely screw that up . But I know that there needs to be controlled impedance on that thing. When you, when you get designs how, how are you looking at these things? Am I telling you, am I telling you that I need controlled impedance on specific traces? Or are you looking at it and knowing that there needs to be controlled impedance here and there because, Oh, I recognize this, that this customer should be doing controlled impedance on these lines.
Ryan:Should tell us. Always be specific. I've seen many boards over the years that don't have controlled impedance requirements on their fabrication drawings. But if you look at the Gerber data you can see that there obviously are controlled impedance traces. And you can do it that way. It works. The downside of it is if, if something happens and those traces don't work, that that weren't specified it can't be a cost to reject that board. You're pretty much accepting it. You have to.
Chris:But because you've got 10 plus years experience doing this, sometimes you can sniff it out. But that, that's not the listeners of the Pick Place podcast are smarter than that. They're gonna tell you that they need controlled impedance on their design. But
Ryan:so here's, here's, here's how the, the listeners should specify their
Chris:yeah. Perfect.
Ryan:We wanna keep it simple. Specify the structure, you know, if it's a coded micro strip and embedded micro strip offset strip line. Let's see what else we have here. Edge coupled and embedded, edge coupled broadside and co plainer strips and co plainer wave guides to ground. Those are the, those are quite a few structures.
Chris:I think I finally understand how my grandmother feels when she listens to the Pick Place podcast, cuz you may as well be speaking Latin to me right now,
Ryan:Well, we will take the time to explain it. We'll start general and get deeper into this if we need to. That's, that's what we're here for. So, you know, you wanna, you wanna make sure you have the, the structure identified and, you know, I, the first one I said was coded micro strip. It, it's often referred to as a single-ended trace. They're commonly attached. It's, it is a single trace. They're commonly attached to antenna,
Chris:Okay.
Ryan:for antenna connections. And they're commonly around 50 oms 75 oms. I've even seen a couple 39 oms, I believe.
Chris:And, and these, these requirements, do they typically come from the chip manufacturer, like Texas instrument tells the designer like, Hey, when you use our chip and you're doing this antenna here's the controlled impedance you need for this specific design. Make sure you tell your fab that you need this.
Ryan:Yeah, most certainly the, the chip data sheets should have the impedance requirements right there. Most of the time when I'm working with Layout designers, they always say, I'm gonna go to the double E to get the impedance requirements. Had one of those this morning. So the electrical engineers often know because they're, they're building the circuit, they're designing the circuit. They, they will know what kind of impedances they need. That information gets passed to the layout engineer and I'm always happy to talk with the layout engineer. I've met some layout engineers kind of like you over the years that they know they need the impedance traces, but they don't know what they need. It is, it's just like another language to. So if we're moving along after we identify, you know, what kind of trace we need, that's what the structure's gonna be. Do we need a differential pair running to a USB or another component? Or do we need a single ended or maybe we have an RF application where we need broadside coupled, or not a broadside couple, but a coded ground wave coded co plainer wave guide. You know, that's a trace where we have, it's like a single ended trace, but on both sides of that trace, we have a ground plane.
Chris:Oh
Ryan:so it, there's another reference there that we have to make sure we calculate for.
Chris:interesting. Cuz my, again, my little pea brain just thinks like, well they're not touching electrically, but I guess because it's like radio waves. They somehow do interact with each other even though they're not physically connected to each other. Because I guess that's the whole point of it, right? Is you're, you're sending out some kind of radio frequency. Now, it doesn't matter that it's not two copper connections, it's somehow gonna have an interface with each other.
Ryan:That's right. That's right. And, and that's a great segue to talk about the differential pairs because we really haven't talked about how they work yet. Kind of like, the co-planning coplanar wave guide, but you now, now you have two traces that are interacting with each other. let's see.
Chris:these traces are gonna come off of some USB chip or whatever it is, and they're going into a port and they need, you know, the, the, the designer of that port and the designer of that chip is saying like, Hey, if you wanna have success here and you wanna have a solid communication, you need to do it this way. And you need to have this controlled impedance in your design to get it to.
Ryan:Yeah. So if, you know, they ha the impedances are gonna have to work. If the impedances don't work, then the circuit's not gonna work. And so the board really is the more than just the foundation to hold the circuit. The board in some circumstances is actually interacting with the circuit to help it work, to do what it needs to do to function properly.
Chris:I think that's what blows my mind about controlled impedance is I've always, I've always thought of a circuit board as basically just like a really fancy cable harness.
Ryan:Yeah.
Chris:all it's doing is just connecting a bunch of things together. But with controlled impedance, it becomes it's own chip. Almost like that's, that's, you know, a gross exaggeration. But it be, all of a sudden, the board itself becomes an integral part of the functionality of the product. It's not just gluing everything together and making the connections. It is actually. Generating something, you're, it's, it's having an impact on the functionality of it, which is
Ryan:Oh yeah. Yeah, most definitely. You know, and, and along those lines, you know, we also build EMI protection into boards with,
Chris:Oh boy.
Ryan:planes, you know, so
Chris:Guess we're gonna have to get you back on to talk about EMI protection.
Ryan:Well,
Melissa:Okay.
Chris:go down that rabbit hole, but, so, okay. If I can, if I can back up for just a second here. You mentioned a specific resistance value for some of these traces, right? You said 50ohm s, 75 ohms, 39ohm s. Again, my little pea brain says, Well, it sounds like it's just a resistor, but I think I'm oversimplifying that it's, it's an impedance value. It's not a resistance value, but that's still measured in ohms somehow
Ryan:is a very good question. Let me explain the difference. A resistor, when you're talking about, you know, the component, uh, of resistor, it, it has no impedance, it just has a value of its resistance. Think of impedance as my college professor always told us to think of impedance as imaginary resistance, which did not help at all.
Chris:No, it doesn't because my professor said the exact same things and I still don't understand it to this day.
Ryan:Yeah. So impedance is the resistance of a copper trace that changes. As the frequency changes.
Chris:Okay. Okay.
Ryan:You can change the frequency of a signal through a regular resistor and the resistance stays the same.
Chris:That's.
Ryan:But when you're talking about copper traces and you, you change the frequency through Copper Trace it's impedance changes.
Chris:Okay, so when you're trying to match 50 oms, you need to know the frequency going through that trace then, is that right?
Ryan:It always helps because it will help with material selection if our customer doesn't know exactly what kind of material they need. I just I was helping a customer yesterday. He said, Hey, I have six gigahertz signals and I have a 50 ohm and a hundred home differential pairs. On the top layer, they're referencing layer two. I'd like to keep the material cheap. Maybe some kind of F R four maybe go. A little if you a little above that if you have to. I said, Okay. So I got out my tools. I'm running with the Polar Field Calculator and Polar Speed Stack. I pull up the data sheets for those materials and when a customer tells me he wants to, to keep it cheap I typically start with f R four.
Chris:I usually run the other way.
Ryan:yeah, Well, because we can get a, you know, we can go, when you start talking about RF designs, you're gonna venture into hydrocarbon, ceramic materials, you know, and, and maybe, god forbid, ptfe.
Chris:Mm-hmm.. Mm-hmm.
Ryan:So, you know, I'll start with an F R four and. We'll just run some calculations and go from there. And, and most of the time, you know, like yesterday, going back to the original question you know, he needed a six gigahertz signal to run on his board. You know, I looked at I don't know if I can mention specific brands, but I pulled up,
Chris:up to you. I, we could care less
Ryan:Well, I, I'm I'm always a big fan of using the, the best tool for the job, but I use a lot of Isola materials because they, they just make good materials. I mean, how else can you say it?
Chris:We, I would say, like when we've talked to various vendors and stuff, Isola comes up a lot. They're, they're big. And if listeners aren't aware, Isola makes for lack of a better term, bare PCB materials that companies like NCAB and their factories will buy and then create products from.
Ryan:Yeah, that's right. And well generally I'll start with, you know, one of their run the mill FR4s s IIsola 370HR and I'll, I'll base it off of that. But the thing is, you know, I look at the data sheet and we going from, you know, a hundred megahertz all the way up to 10 gigahertz and so I chose the 10 gigahertz material. So that will cover us all the way up to 10 gigahertz. I calculate the impedances for that material. For, in those specific traces, I give my customer a specific stack up and they should be able to take that stack up and put it into their design and use.
Chris:Oh, interesting. So, so then that, I guess that means if you, if you have this requirement, you should probably have this conversation fairly early in the design process because it may influence your layout and, and you know, you don't want to, you don't like, you don't wanna have your layout finished and then go to NCAB and say, Hey, we need controlled impedance here. Is this something that you should start the conversation with a little bit earlier to know what your stack's gonna be?
Ryan:Yes, most definitely. I love to work with customers up front during the design phase. It's, it's one of my favorite things to do because not only do I get to, you know, I love to help people. So not only do I get to do that and, and help with their design, but it makes it so much easier when it comes to manufacturing. You know, we get all that stuff worked out up front. We get the Gerber data. Good. You know, it gets to the quote phase and there are a lot less questions that have to be asked and answered.
Chris:Mm-hmm. We, we love to be able to get into the conversation with customers early on the design phase, if possible, to help with DFM design for manufacturability because it just makes everything go so much smoother rather than trying to, you know, like, Oh, I see what you're trying to do here. We've already got the board's fab. We'll, we'll trim this off a little bit and make it fit, and oh, I gotta build 150 of these. And then, and then inevitably they, they go and place their second order for 1500 without making any of the changes. You're like, Oh, no, I,
Ryan:Oh yeah,
Chris:So, yeah, it's nice to get those conversations started nice and early.
Ryan:Yeah. Yeah, it definitely helps. And you know, that stuff happens.
Chris:I didn't realize that it was valuable to have the conversation about controlled impedance early on in the design phase, because so often, well, I mean, so often when, when I'm getting involved, usually the design is more or less done and they say, Hey you know, here's my Gerber data, here's my fab drawings and everything. Oh, and here's the information about my controlled impedance. And that's the first time that our fab is hearing about this controlled impedance. Is that u, does that usually work out okay? Like if the design engineer is familiar with what they're doing, or, or is it sometimes a little bit painful and you guys have to, you know, make it work, so to speak?
Ryan:It goes, it, I've seen it go both ways. You know, let's go talk about the design engineer that knows what they're doing. I've, I've met many men and women out there that design controlled impedances and they don't need my help. And that's, that's great. That's okay. That's the goal, right? But then there are some, like I spoke about before that do need a little bit more help. And most of them are just college graduates. They're just starting out. And I
Chris:think that's most of our listeners, to be honest with you, Ryan , I think it's mostly junior engineers who are trying to get up to speed on these things.
Ryan:Those are my favorite engineers to help. And we'll talk a little bit more about that later,
Chris:Sure,
Ryan:but they're always, they're always open to suggestion. I, I guess a nice way to put it is they haven't been jaded
Chris:Jaded.
Ryan:by certain aspects of this industry yet.
Chris:Yeah. Yeah.
Ryan:So yeah, it's always best if you have any circuit board questions to get in contact with us early in the design phase. As far as FAEs go, like me we have several on my team. That's what we're here for. It's, it's my daily goal to get in front of those engineers during the design phase. That's all my job is.
Chris:That's beautiful. So I will say then for, for Worthington, if you're you know, working on a design that has controlled impedance by all means get in touch with us. We'll, we'll, we'll give you the connections directly to our suppliers. Usually it's probably gonna be ncaa And you can just have those conversations. Like we we're not gonna middleman at all. I would say for Melissa's group, she's got it even better. She's got in-house cam engineers and design engineers that work directly for CircuitHub. So, you can have a direct connection with those guys, whereas I gotta pawn you off to guys like Ryan. I can't I can't help you directly. But so I will say now when it comes to, like, you, you were talking earlier about the various I want, I want to get more back to specifics. The various resistance values. Should I call it a resistance value or should I call it an impeded, I should call it an impedance value.
Ryan:Yeah.
Chris:keep hearing ohms and it, my head keeps going to resistance.
Ryan:Yep. Yep. It, it's it took me a while too. First thing I had to learn how to do was spell impedance with an M and not an N then
Chris:There you go.
Ryan:Yeah, I, I'm being sarcastic, but the guy
Chris:think there's an n at the end of the, at the end of the, the word.
Ryan:Yeah. Yeah. The guy that trained me you know, he showed me the basics and we stayed on it for years and years. It, it wasn't something that we just did one class on and he, he let me go because controlled impedance traces are that important. They have to work.
Chris:And you're using, it sounds like probably pretty fancy software to do these calculations. I don't think it's something you can download for free. I'm sure it's licensed and, and expensive. Yeah.
Ryan:yes, it both, it's, I'm using the, the polar speed stack and the polar field solver. I've used the, the field solver for many years. The speed stack I have recently got into in the past year. I've got that up and running for our customers and they, they love it cuz if you do need to specify controlled impedances, we can have a conversation and I can learn about your board and you needs, and, you know, the same day I can have a stackup, a a at least a preliminary stackup in your inbox.
Chris:Nice.
Ryan:Whereas, you know, if, if you go to a factory direct you're gonna, you're gonna wait for their time. And if you're going there just to get a stack up and you don't have a po they're not gonna put you on the top of the
Chris:that's right. I assume that the people who are developing these tools, they they're either, they either have scientists or they're taking research that scientists have performed to evaluate all these things, to come up with all these calculations. There, there's probably been a tremendous amount of work that's gone into you being able to plug in some numbers and get an output.
Ryan:most certainly. I've met with the people at Polar.
Chris:Oh, cool.
Ryan:and I mean, if you look at the controlled impedance specification, it gives you the formulas, these ungodly formulas. So the, the good people at Polar have they have scientists, they have engineers and that understand those calculations and everything that goes into it. And they make it easy for us. They put it in this nice user interface. But probably, you know, those tools are great to have, but if, if I don't have a data sheet out for that specific material then it's not gonna be the best product as possible. So I'm always pulling out the data sheet. That's, that's a really good tool.
Chris:And you're talking about the data sheet for like Isola or whatever it is, whatever material it is that you're trying to build with?
Ryan:Yeah. For the material we're building with, whether it's Isola vtech or Rogers or we get up into the PTFEs. The, the most important value we're trying to get from the data sheet is the dialectric constant. And it'll be shown on the data sheet as uppercase d lowercase a, you know, subscript K
Chris:Mm-hmm.. Yeah. We we, we publish on our website kind of the standard run of the mill bread and butter stack ups, and a lot of our customers use those stack ups as a, you know, I wanna get by with the cheapest material possible, so I wanna stick with the standard stack up that these guys are churning out every single day in their factory. And I, I believe some of our customers are somehow doing their own calculations for controlled impedance. Do you see that much, or is that kind of a rarity? Do you, do you guys normally end up doing that?
Ryan:I do see it. You know, I, I hear, I can't talk intelligently about it, but I hear about the Saturn PCB calculator and some other free online calculators. And you know, just last week I had an engineer that he did his own work and just wanted me to verify his work with a different calculator.
Chris:Yeah.
Ryan:He nailed it. You know, it was right on, and we were both happy about it.
Chris:Yeah, I would be too.
Ryan:It doesn't get any easier than that for me.
Chris:So how do, Okay, so that's how we, we, we get the design right, is we work with engineers like Ryan and or, or we get the right software that works for us and we get the stack up and everything. But how does, Okay, so now we've handed it off, right? Ryan's done his work, he's done his calculations that the, the design is locked in. We hand it off to the fab. How, how is the fab now? Like controlling the quality, verifying that that's, that's gonna come out right? Because I know that we get a lot of times our suppliers will put like an impedance, forgive me if I'm using the wrong terminology, but like we call it an impedance coupon, they'll put on the breakaway rails for us. So like, we have these breakaway rails so that we can, we can automate the assembly process and solder everything together and that usually gets thrown away. Well, I, I believe the fabs find that as convenient location to put a impedance. Coupon to. Is that to control their process or is that something that they do so that that, that their customers can confirm that their impedance is correct. I'm curious about that whole quality control process.
Ryan:Okay, well I'll start there. Yeah. The impedance coupon is exactly that. It's for testing to ensure that the impedance traces were manufactured correctly. And one of the common tests is time domain reflectometry. It's it's an extra cost, an extra process that your manufacturer will have. But they, yeah, after, you know, somewhere between final inspection and shipping, they will connect those impedance coupons to that machine and just verify that the impedance traces on the board work.
Chris:So they, they connect these coupons, I assume they're connecting the coupons to this reflector machine. I'm gonna call it reflective machine.
Ryan:Yeah, we'll, we'll just call it a, a TDR test for
Chris:TDR test. Thank you.
Ryan:So, so if, if you have to have your impedance traces verified ask for a TDR test. But the, the TDR coupons are designed during the, Let's go back a little bit.
Chris:Sure, please,
Ryan:go back after everything's done, you, you have your final design stack up. Impedance traces are all good. You send everything to the factory. The first thing the factory's gonna do is they're gonna pre engineer that board. So they're gonna have an engineer go through the specifications that came with it, the Gerber data, and they're gonna make sure everything meets specifications. One of the final things that we will do is pull out that stack up check. All of the traces on it. If there's a impedance trace on the stack up, then we wanna go find it on the Gerber data verify that's there. You'd be surprised how many times it's not there.
Chris:Okay.
Ryan:That's, that's a different story. We'll, we'll maybe we'll talk a little
Chris:What, What do you mean the traces itself aren't there? Or you're, you're saying the specification for the, for the traces
Ryan:Well, most likely what has happened is an engineer has copied and pasted, you know, specifications from one drawing to the other, didn't catch the impedance stuff that doesn't need to be on the new board, and they just left it there.
Chris:yeah, yeah, yeah. Yeah.
Ryan:So, yeah. What, what we'll do is, we'll, we'll make sure the traces are there. And then we'll verify that, you know, the, the wits and the spaces, if you know if it's a differential pair, do they have the proper spaces between traces? We'll make sure the trace widths are you know, if they're supposed to be five mils, then we'll make sure they're five mils. Another thing we'll do is check the the planes to see, because impedance traces, we haven't talked about this yet, impedance traces have to have a reference. So they're gonna reference either one or more planes. And we'll make sure that those planes are where they're supposed to be in the Gerber data. Actually I did some work for an engineer once designed a stack up for him the way he wanted it, and when he gave us his design back, he left a plane out. From where it was supposed to be for referencing the impedances, and God forbid this ever happens, but it does. Nobody caught it from, from him all the way out to final inspection. And he got the boards back and they didn't work. And we went back to see what was wrong and he, he called me, he's like, I'm so sorry. I said, No, I'm so sorry. We should have caught this too. You know, things like that happen sometimes. But we wanna make sure the planes are in the correct area. Then we're gonna go through and you know, if we have a stack up, we can just pull up the field solver and input all the values from the stack up. And everything works out. It goes on to the next process.
Chris:Okay.
Ryan:begin, you know, they pull the material because of the stack up. They know what material they need. They start cleaning it and start manufacturing the board.
Chris:Yep. And, and, and so. You, you mentioned something earlier that I wanted to ask you about. You mentioned that, you know, if, if it's defined as five mil, you're gonna make sure it's five mil, but you guys have like etch factors and these sorts of things. So you, so if, if, do you do, do you have to do anything different for a trace that has controlled impedance as you would etch factor on any other trace when you're, when you're calculating how, how you're gonna fabricate these things?
Ryan:We look at them differently. That's, there's two, there's different ways to approach that. Right? Here's, here's what we want. Let's talk a little bit about etch factor. You know, we're. Have these copper clads and foils, we're gonna print the circuit on and we're gonna send it through a machine that's gonna etch the copper away. There's a certain rate that that material goes through the machine, and that's based on the copper wave. And that's based on the space, on the electrical space of the board.
Chris:Mm-hmm.
Ryan:it's one of those things, how much etch factor, you know, are they gonna add in pre-engineering to make sure that when this copper clad is etched, how much etch factor are they gonna need to make sure the traces do come out at five mills? Well that's factory dependent. I've, I've seen factories add a lot and a little, and it depends on, How, you know, if it's a production factory, they're gonna add a lot. Cuz they're just running through those, running those boards through there constantly. They're, they're not babysitting the machines. They don't have time for that. But if, if you can stand there. Yeah, yeah. The computer, it's all computer controlled, but in, in smaller factories. Mostly in the us you know, it is not as advanced machinery say. And so they're gonna add edge compensation in a different way. But let's, let's, let's start with a baseline. So let's say we have you know, we're etching through half ounce copper. We're going to. Be there at the machine, everything is flowing smoothly. We're probably gonna add a, maybe about a half a mill of edge compensation to those traces. So when they, the, everything's gonna be beefed up by a half a mill before it goes into the machine. When it comes outta the etcher, it's gonna be within the 20% cutoff that the IPC allows for.
Chris:Okay.
Ryan:So the more etch factor, the more etch compensation that we add, and the better we etch it, the more accurate the traces are gonna be. So, getting back to your original question, now, let's compare a regular trace to a controlled impedance trace. You know, the, the I P C allows for a 20% tolerance there when that trace comes out. After plating, you
Chris:Sure.
Ryan:so if it's a regular trace, and let's say it's, let's go far, let's say it's 21% under what it should be, Um, you still may get the performance. You know, if you have low voltage, you, you're still maybe gonna get to the current performance you need. Or let's say it's, it's under etched and it's 21% over is still gonna perform most likely the way it needs to perform. But an impedance trace most likely won't.
Chris:Interesting.
Ryan:if, you know, you talk about. A 50 m impedance and most impedance tolerances are 10%.
Chris:Plus
Ryan:can only go, yeah, you can, you only got 50, 55 ohms on the top end and what, 45 oms on the bottom
Chris:Mm-hmm.
Ryan:If, if you changed the trace with by a quarter of a mill you could certainly not meet the, you know, the board would certainly not meet the speced, the impedance specification.
Chris:Oh, wow.
Ryan:you know, answering questions like this are subjective because people wanna know, Okay, well what will it be? There's no better answer than just getting into the field solver and crunching the numbers,
Chris:Yep, yep.
Ryan:And I can tell you, okay, what happens if we're gonna be out this way and out that way? But that's really where. It really comes down to the manufacturer there and how much electrical space is on the board. That's what's in the designer's control. I always say the more electrical space we have to add full edge compensation, the better. If we can add full edge compensation, then the impedance traces will come out beautifully the way they're supposed to. But if this is one of those beast boards, you know, it's got blind vias buried vs. We have to add a whole lot of extra copper plating because of via fill. Then you know that extra copper is something that we have to add even more etch compensation to to etch through to make sure the impedance traces are as close to the targets as possible.
Chris:this is the common theme of the Pick Place podcast is make it bigger. Don't , don't get fancy with it, you know? Yeah. You know, give us, give us the room we need to get the, to get the work done. But now, now that I have this fancy new equipment, I'm gonna start to tell people the opposite. Make it as small as possible, because then I'm the only one who can build it.
Ryan:Yeah. Be careful what you wish for. Chris
Chris:exactly. No, true. So true story. We, we have never in the history of our company, not on purpose, it's not like we were avoiding it, but we had never in the history of our company had a design come in with a Imperial oh 1 0 0 5, which I believe is a metric metric 4 0 2 component, you know, capacitor, resistor. And literally like the week after we installed these machines, we got a request to, you know, a request for quote with a 1 0 5 s of like, what are the chances that we get this new equipment?
Ryan:Buckle up
Chris:them. Yeah. Anyway, so I. I think, I think I'm still a little fuzzy here on understanding why, how do I put this? Well let me, let me explain my view of it and then, and then that might inform your answer. So, when, when we get these impedance coupons from our fabricators, sometimes they come as a totally separate thing. It's like, it's like just included in, along in, in the package along with the PCBs, but sometimes it is in that breakaway rail, like I mentioned. Why, why do fabs. Do that, Do they, Is it because people are requesting these things? Is it because it's their way of saying, Hey, look, here's the proof that we did it right each and every time. Is anybody doing anything with these coupons? Like, I'm not doing anything. I'm just passing 'em along to my customer. Are are your, are the end users, not the end users of the boards, but the, the designers of these boards, are they using those coupons for anything? Or is this just like a good faith sort of thing, saying, Yep. See, proved it, we, we did it right.
Ryan:Yeah, well, it's both actually. You know, sometimes customers do request impedance coupons and that's fine. We just send them along. But in your case where it's on the rail, right,
Chris:Yeah
Ryan:we don't wanna remove that rail before assembly. So if the impedance coupons are built into the rail they're gonna go along for the.
Chris:that's right.
Ryan:But
Chris:they'll maybe be impacted by all the clamps and heat and everything. I've put them through
Ryan:most certainly. And then you can, then you can test them again. If you have the equipment and the knowledge, you can, you know, you can test them after that.
Chris:yeah. So I wonder if maybe larger shops that aren't doing such low volume, maybe they're, maybe they are doing some some testing on, on those sorts of things that we do some testing, but most of the testing that we do is mostly kinda like functional testing. So, you know, power it up and does it connect to wifi and does it get a cellular signal, and that kind of stuff. You know, we're not, we're not really putting them into the would you call it a TDR machine? Is that
Ryan:Yeah. Time domain, reflectometry tests and tdr. Most of the boards that I've seen that require that test are your military and aerospace type board.
Chris:sure. That makes sense.
Ryan:you know, because they, that TDR test is just one of a bazillion tests that those boards need. You know, if, if your toaster needs to connect to wifi or usb it most likely it, it's not gonna need a TDR test.
Chris:Yeah. Yeah, yeah, yeah.
Ryan:and, and another way it, you can do, you know, if you are working with your manufacturer, you can say, Hey, can you send me your impedance calculations? I wanna see, you know, the most, most fabs I know they're gonna produce an in-house stack up based on the customer stack up. And as part of that process, they're gonna calculate the impedances. So it is, if you have a relationship with them, you can call 'em up, say, Hey, can I have your, a copy of your stack up in the impedance calculations? And many times customers will do that in lieu of the extra cost associated with TDR testing.
Chris:Okay. Okay. Well that's good to know. Well, I'll tell you what, I think this is a good place to, to ask if you have any specific tips or recommendations for designers. If they know that their design is gonna require controlled impedance, you know, what are the, Hey look, you gotta make sure you do these two or three things and your life will be good if, if you're, if you're, if you're aware of this and you take the time to think about it everything will go so much smoother for you. You know, I guess, I guess what I'm getting at is like, what are people assuming that they shouldn't assume and, and what information do you get that makes just everything go so much smoother when you have a controlled impedance circuit board, or excuse me, a circuit board with controlled impedance.
Ryan:Oddly, I think the most important thing to have in the design is enough room for complete etch compensation. You can have the best design stack up in the world with the best design impedance traces, and if the fabricator can't add the necessary amount of etch compensation to etch those traces to what they need to be, then none of it's gonna work.
Chris:Mm. Mm-hmm.. Mm-hmm.
Ryan:you know, that begs the question, well, how much etch factor do we add?
Chris:I was gonna ask
Ryan:how much are we gonna add? It's, it's gonna be dependent on different from one manufacturer to the other. But that's, that's where the engineers can work with the designers and say, Okay you know, like, well, let's just go for an example. An engineer comes to me. I'm gonna ask for the fabrication, drawing, and the Gerber data. I wanna look at the Gerber data, see how much space we have, What other kind of technology does a board require that's gonna influence that manufacturing, You know, are we gonna, if we're starting with half ounce copper, are we gonna etch through half ounce copper or are we gonna add a little bit more and then have to etch through it?
Chris:Okay.
Ryan:so I think the number one key thing is develop a relationship with your board supplier and ask those questions upfront during the design phase. And, you know, at NCAB group, we're always happy to help. We, we have a staff of engineers that do just that.
Chris:Brilliant. You, you know what, you are mirroring everything. We always say, we always say the same thing. It's give us enough room and have a conversation
Ryan:Yeah. Yeah. That's it.
Chris:you know, And, and by having the conversation and by sending screenshots and dimensions and things like that you, you, you just save yourself so much stress and frustration and time and effort by, by solving those things early on and not running into those issues later on. So yeah. Great. Excellent, excellent advice. Before we get into my favorite part of the show, the pet peeve of the week, is there anything further you'd like to talk about when it comes to control impedance or anything you wanna share that we didn't get a chance to touch on?
Ryan:You know, there, there are many other things that we could talk about. We could probably have you know, another one hour conversation.
Chris:I believe it.
Ryan:About, you know, how do we change control impedance traces and what influences them, You know, the dialectric and the stack ups, we didn't get into any of the design portion.
Chris:Well then we'll have to have a second episode,
Ryan:maybe we can do another episode and just focus on you know, we can go into detail about what the engineers are gonna do when it gets to the factory and how that's gonna influence manufacturing the board.
Melissa:Mm-hmm.
Chris:let's, let's, let's pencil that in then, Ryan. Let's let's make sure to get that scheduled and have you back and and talk about those specific things. That'd be great. Especially you know, now we covered kind of the, now that my friend knows what controlled impedance is yeah , we can, we can get more into the nitty gritty. I think that'd be great. Because we do try to keep these episodes to a, at least a somewhat reasonable length. We've, we've definitely gone long on some of them, and this is one of our longer ones. But by all means, let's, let's make sure that happens. I think, I think it'll be great. Especially, I think what'll likely happen is we'll release this episode and then we'll have follow up questions. We'll have listeners that write in and say, Hey, I wanna know about this and I wanna know about that. And we can include those in our future conversations. So if you're listening to this episode and you, you didn't get your question answered about controlled impedance because my friend didn't ask the right questions by all means right into us and let us know and we'll try to cover'em in the future episode.
Ryan:yeah, most certainly. I'm, I'm always here to try to help and answer those questions and if I, you know, I don't always have the answer, but I'll find it.
Melissa:Can't
Chris:That's it. That's it. We will find it. If the answer does exist, we'll find it. All right, Ryan, let's do it. Let's do pet peeve for the week. What? What is your pet peeve that you would love to share with us? I'm excited for it. Lay it on me.
Ryan:you know, my, I I think my number one pet peeve are, are you know, and I'm trying to be tactful. I'm not very good at tact, but, you know, I've done this for 11 years and I've, like I said, worked with engineers from all walks of life. And the engineers that need help are one thing. The engineers that don't need help are another thing, but the engineers that just don't care.
Melissa:Hmm.
Ryan:Those, they, they still call me because they have to work with me. But I can tell that you don't care. And I see it in your Gerber data and on your fab drawings. You're not sneaking past me. I'm still going to, I'm still gonna do my best to help those engineers out.
Chris:You know what's funny? I don't think it's ever crossed my mind before that an engineer wouldn't care.
Melissa:Hmm.
Chris:But now that I think of it, when I look at some designs and I think what this board is 10 inches by 10 inches. Why do you have oh 2 0 1 s on here? It's cause you don't care.
Ryan:Yeah. Yeah. And you know what? There are a lot of, in the electronics industry, there are a lot of black boxes out there. So more power to the black boxes. I love those that, that concept. But, you know, in my previous life I was active duty for 10 years and it got to a point where I was ready to go. It wasn't for me anymore. And
Melissa:Mm-hmm.
Ryan:I understand engineering world is, is like that too. It's not for everybody. But you know, we're all here to end this together, so let's let's work well together.
Chris:I, I don't understand. I think maybe, perhaps the reason I never expected an engineer not to care is because I can't even fathom not caring about something I'm working on. Like I.
Melissa:Mm-hmm.
Chris:Like, it's, it's, it's a broken concept in my mind because I, if I'm gonna do something, I'm gonna do it to the absolute best possible ability that I can. I, you know, it's like, I don't care if it's like washing dishes or, or, you know, helping a customer with controlled impedance. I'm just gonna put all my effort into it to do it. Right. The concept of not caring is so foreign to me, But I, But you're right there. Prob that probably is a real thing that happens.
Ryan:Yeah. You know, and, and I'm probably being a little too abrasive there. I don't want the, your audience to, to get the wrong idea that I work with a lot of engineers that don't care because I don't it, it's very rare, but it, it's always there.
Chris:I guarantee you, Ryan, I guarantee you. If somebody is listening to the Pick Place podcast, it's because they care
Ryan:Yeah.
Chris:I guarantee you.
Ryan:and that's the kind of audience that I love to cater to.
Chris:Sure. Yeah. Yeah. This is a, we, we gotta, we have a great group of listeners. I, I will say the feedback we get from them is just, I love reading it. I love talking to them. We've, we've actually earned some customers through this thing, which is, you know, it's secretly, of course, we, we want this to be an avenue where we, where we get business, of course, but we also wanted to create something to teach people. But when we, when we do have people that come to us that, Hey, I listen to the podcast, say hi to Chris and Melissa, blah, blah, blah. Inevitably those are the best designs. Like, inevitably they're the easiest things to build. They're, it's like, Oh, it's cuz they care. They're, they're listening and they want to do a great job. And that
Ryan:listened and did exactly what you said. And yeah, I love helping those engineers. I'll, I'll work late to help those engineers if I need to,
Chris:Amen to that. Amen to that. I, there are specific customers that I know have, have listened to the show and have reached out to me. And I, I do, I I have a soft spot for 'em. I bend over backwards for 'em. I, I, I totally go outta my way. I work 'til mid, like midnight, like you say to help those folks out because when, when you feel their passion and you feel their energy, it, it's just infectious and you wanna give it right back. So, Amen to that, Ryan. I think that's a great way to wrap it up. I really, really appreciate your time. I'm looking forward to having you back. I'm excited to cover future topics
Ryan:Well, I appreciate you giving me the time to come on here and I appreciate the plug for NCAB Group.
Chris:Yeah, absolutely. Yeah. Gotta keep this whole thing rolling. If you're listening to this and you have further questions for Ryan I can, I can absolutely put you in touch with him or, but I would prefer if you wrote to us contact@pickplacepodcast.com so that we can, we can roll it into a future episode as well. And perhaps you know, obviously if you want to get in touch directly with him, cuz you have a specific need for it, by all means, we'll, we'll make that connection. As always, you can reach us at w Assembly and at CircuitHub and and keep those show suggestions rolling in. We love, we'd love to hear 'em. We really appreciate it and you've definitely been telling your friends about it because our numbers are growing despite the fact that we, we have not published an episode for six weeks. We still get big downloads, which is surprising.
Melissa:Yeah. I've been expecting them to go down, but they actually haven't so
Chris:They haven't, which is the strangest thing.. It's the bots. It's all the bots I'm hiring to to download. Yeah, that's gotta be
Melissa:On that note, thanks for listening to the Pick Place podcast. If you like what you heard, consider following us in your favorite podcast app, and please leave us a review on Apple Podcast or wherever you get your podcast from.
Chris:Thank you again, Ryan and Melissa. Thanks everybody.
Ryan:You're welcome.