MacroFab Engineering Podcast #191
The US Mint Denver produces 30 million coins a day. Denes, the tooling department manager, discusses with us how production at this scale functions.
Stephen is on the hunt for the next step in his electrical engineering career and shares the shifts in the industry and what employers are looking for.
Relay manufactures hate this one simple trick that makes your “sealed” relays last longer! Except TE connectivity who has an note about this relay feature.
PCB Manufacturer Reviews
Special thanks to whixr over at Tymkrs for the intro and outro!
Welcome to the macro fab engineering podcast. I'm your guest Daniel Hinch.
And we're your hosts Parker,
Dolman and Steven Craig.
This is episode 191
Daniel Hinch is the founder of Rheingold. Heavy and manufacturing reports. Dan is passionate about education and works every day to demystify technology and teach how to measure the world around you. Rheingold heavy and manufacturing reports is Dan's effort to provide the community materials he wished he had when he was first getting started in electronics and technology. Dan also drove a forklift for a living and suggested you should give it a try.
So thank you, Dan, for coming on to our podcast.
Absolutely. Thank you for having me.
And before we jump into this, I want to hear a forklift story.
So one of my one of my first jobs out of high school, was working for a power transmission company in Benicia, California. And by power transmission, I don't mean like power transmissions in cars. It's a power transmission like gear, motors, electric motors, worm gear drives and stuff like that. And we sold these things to pulp and paper mills to there used to be something called a new United Motors plant up in San Jose. Any place that needed big electric motors or gear drives, or to move a conveyor belt or something like that, we were selling into those companies. And so this is a good story. We had a UPS driver named Ross, and I remember what Ross his last name was. But I very specifically remember that Ross had on the side of his van, a little plaque that said 10 years safe driving. And, and one of the things that Ross used to do is that he would pull up to the top of our loading dock, put his van in neutral, and then just let it roll all the way back until it crashed into the loading dock at the end. And you hear this huge crash and then you'd roll up the door and you know, oh, yeah, Ross is here. So So Ross comes in. He did this several times. And when I would go to lunch, he would get on the forklift because we had a locked warehouse. So you know, it was nobody going to steal the forklift, he would get on the forklift, drive it over and pick up my desk in the warehouse and drive it and leave it on the top of the pallet racking. And then take the key out of the forklift and go do deliveries. And I would come back from lunch and have to do all of my paperwork for the shipping in the afternoon. So I'd have to climb to the top of the pallet racking stand there doing filling out all these bills of lading for probably an A like 150 200 200 orders, and then climb down, try and do all as much of the fulfillment as I could, considering that, you know, some of these, these motors are weighing, you know, upwards of two or three tonnes waiting for us to come back with that, that that forklift key so I could actually move that thing you know, from from one end of the warehouse to the other. It was also great, because we had a really well polished concrete floor back there that my my supervisor, Randy Metz would, would get on the forklift and I would hang on to the side of it in my rolling chair, and he would accelerator as fast as he could. And then when we hit top speed, you'd let go of the forklift and you just go spinning, you know, at about probably 1520 miles an hour across the warehouse floor. And the fact that those things have the rear wheel steering is just cool. You know that it sits there and spins on that inside front wheel. So you can crank that steering wheel all the way over to one side or the other. Put just a little bit of pressure on the accelerator and then hop off and then come back five minutes later and it's still sitting there corkscrewing in the middle of the warehouse.
So it stories like this, why logistics at McWrap still don't have a forklift?
Oh, yeah, those things are those things are there a lot of fun, especially when you pick up a like, I think we had an 1800 horsepower, totally enclosed fan cooled electric motor from Toshiba. And that thing weighed probably within, I'd say 15 to 20 pounds of the weight of the counterweight on the back of the forklift. So you lift it up and you'd basically just sit there and rock on your front wheels. As you're sitting there trying to get that thing into the back of a of a you know, a little pup
freight truck. You get all the interns to sit on the back of the forklift.
Yeah, exactly. So yeah, forklifts they are, they are a ton of fun. At one point, I hope to have some reason to own one.
So Dan, talk more about you. So who are you?
So I am Dan Hinch. I come from a background of forklift driving and electric motor repair. Although that translated very quickly into working in information technologies for Jesus probably almost close to 20 years in, in everything from financial sector to work for a startup in Silicon Valley for a while, but mostly in financial services, particularly down here in Los Angeles. And it took me 20 years to discover that I was awful at that job. So finally discovering that maybe
you discovered it, or like your supervisor, slash boss, or
been my or a succession of employers that suggested that perhaps I wasn't the best fit for, for, for telling people, you know, have you tried turning it off and on again. So. So, as it came to pass, I found myself between jobs and also happen to have had the forethought this time to actually save up a bit of money so that I didn't immediately need to go find another job and decided, you know, I've been, I spend all of my time at work, thinking about what I'm going to do with electronics, when I come home, I'm going to I'm going to sit there and I'm going to futz with this circuit, or I'm going to try and try and read through this textbook that I got from the library or something that I bought off of Amazon, or I'm going to build a kit. Why don't I try doing that for eight hours a day instead, and see if maybe, I can find some way to get people to pay me to do that. And at the time, I was also taking Chris Gamble's contextual electronics course, which was talking about PCB design. And up until that point, it had all been perfboard and bread boards and and and you know, hand soldering, you know, almost beam robotic circuits together just component to component, it never occurred to me that hey, maybe, you know, doing doing e CAD was something that I would be capable of doing. And his course on, you know, how to learn chi CAD and how to, to take, you know, your schematic capture and actually translate that into an actual circuit board, kinda came up at the same time that that, that OSH Park was really starting to come to the fore. And the idea that you could actually just get a circuit board made for a reasonable amount of money, not even just a reasonable amount of money, but but a inexpensive amount of money was, was groundbreaking, for me, at least personally, that I could design a circuit and actually have a circuit board delivered to my house within 15 days, for you know, 20 bucks, was, was inconceivable until I realized that it was in fact conceivable. And then I started to have a lot of ideas for things that I had struggled with over the years that there wasn't any real decent quality content for and because of that, I wanted to fix that. So anybody that's getting into electronics, anybody that's interested in electronics immediately discovers that the basics are covered. And not only the basic is covered, the basics are covered five ways from Sunday, if you can, if you don't like the way one person is covering the topic, you just can switch to another person because they'll have a different tone, they'll have a different way of presenting it, they'll use different analogies or whatever. And you'll eventually find somebody that is speaking your language when it comes to the way that you learn that information. But man as soon as you move past the basics into even the intermediate the quantity just the physical quantity of material starts to die away. And before you know it you are you are back into people that are describing everything with derivatives and integrals and feminine theorem. And and most people just sort of switch off at that point. So I wanted to sort of attack that next level of information, which is why I decided what I was going to do was cover sort of sensor signaling not exactly high speed signaling but but I squared C and SPI and really break it down not from the here's how you use this library. But let's let's analyze the the timing for the signals. Let's see what it looks like on the oscilloscope. Let's understand how you can add Actually troubleshoot this so that so that you are no longer beholden to what somebody's library is telling you it's capable of doing. But you're actually capable of, of modifying that library or writing your own library, you're actually able of understanding what each individual chip is doing and how you can do that most effectively. So that was my that was the first product that I put together was the I squared C and SPI Education Board, still available for sale at wrangled. heavy.com. And, and I've and I've sort of grown from there to the point where, where I now have that and and as of I guess, early last week, I started manufacturing reports. That was
a that was a hell of an intro. And it's a story, there was like four segways that could go off and do a bunch of other things there. I love it. But but I'm actually really curious to hear a bit about Rheingold heavy. And what that is and what, what's behind that before we jump into manufacturing reports.
Sure. So wrangled heavy is the venue that I hit upon to actually communicate this information that I wanted to present to the world, which is the website that I run. And I had at the time, I figured there were there were two ways of going about it, I could either sell the content, or I could sell product, retail product. And the idea of selling the content itself that written content. I mean, it didn't, I don't want to say like struck me as being evil. But I had a choice where I could either put it behind a paywall, or I could make this information available to anybody who needed it. And if there was anybody out there like me, I would want to be able to get to that information and try to understand it, the thing that I that I hit on was to tie the content that I'd written to the to as a as a hierarchical education system tied to the board that I had designed. So there is a lot of really basic theory that is universally applicable across, you know, doesn't matter whether you're using, you know, a ti Launchpad, or you're using an Arduino spy signaling a spy signaling a spy signaling doesn't matter. But if you are, but when you start getting into the how do I actually use this, then I wind up talking about this specific chip that I chose to put on that board. And I talked about how you read that data sheet, how you read that timing diagram, or you understand the registers inside that data sheet. Because you can't really do that universally, you kind of got to pick something and stick
with it, you have to have an a working example. Exactly. And what
I really tried to focus on doing is delivering the content in a hierarchical fashion. So it's, so you can just jump in, like I mean, the probably the number one page on my site is still how to choose the size of an iceberg. The pull up resistor, like nobody actually starts at the beginning. Unless you own the board, nobody starts at the beginning and just reads their way through to I squared C pull up resistor, there's a lot of people out there that are trying to understand how exactly big is this thing supposed to be. And I've got a page that describes it in in exquisite detail. But the chips that I chose, and the way that I presented that material and the way I laid out the board. So one of the the ideas is that I want this to be a hierarchical system. So if if you have bought one of these boards, the and you follow the content that goes with the board, what you actually start out doing is manually clocking seven, five or some for hc 595, shift register, you you physically manually press buttons to affect the clock, the latch and the data. So that the all of that stuff is happening at a human conceivable speed. And then you step that up to I squared C and then you go from I squared C to SPI slowly increasing in speed, increasing in complexity of the chip in increasing in complexity of the register maps, as you're going and understanding how that signaling is working, how the timing diagrams are working and most importantly, how to troubleshoot that stuff. As you're as you're going into more and more complex systems, more and more complex sensors, more and more complex micros
instead of just jumping into 20 megahertz and just slamming you know, whatever code and hoping it works.
Yeah, exactly. And mind you there there are absolutely people that that simply want to buy component A and component B, and make the to work by following a recipe. And as soon as it's working, it's doing whatever they want it to do. And that is fantastic. If if that's all you want to do, I am, I am not going to stand in your way to doing that. I'm not going to tell you that's wrong. But I know that there are people like me that that need to see they've got this compulsion to understand what's happening behind the scenes. And I personally find it difficult to use somebody else's library, I want to either write that library myself, or at least be able to pull their library apart and understand what it's doing, before I incorporate that into whatever it is that I'm building,
this is gonna be a side tangent, but it's on that is, that's what I get, I don't like about all these new IDs for microcontrollers, but Atmel like microchip are doing, we're like, there was like, Atmel start. And I think Microsoft's got the same thing where like, they have like, eight levels of abstraction between you and the metal, right? Like, and you use the example, an example is seriously like, start it calls an IT function, and then a while loop, and it's like, that doesn't help me make this thing work in my own products,
that sort of takes the fun out of it, you know, like, like, overcoming the the obstacles and the, I guess, deficiencies that a micro has, that's, that's like 90% of the fun, you know,
that's, and that's the thing is that I don't look at those deficiencies, as deficiencies, I look at those as efficiencies as I have, I have removed all of those abstractions out. So now, I am able to actually flip the very single specific bit that I want to flip in order to get it to achieve a specific result. But I'm also not trying to, you know, control satellites. You know, I'm trying to read data out of a sensor. So I would prefer to do that as efficiently as possible. And honestly, the thing that I, I find probably like, the biggest challenge, like the big puzzle in any project is power management. Like that is that's when you're getting down into the nitty gritty is, how can I get it to draw the lowest amount of current that I possibly can? How can I squeeze the most current efficiency out of this particular design? When it's supposed to not be doing something? How close can I get that thing to being turned off, and then still allow it to respond when I want it to respond?
So then, so how did wrangled heavy, like, evolve into this manufacturing reports? And what is manufacturing reports?
So it's, it's really simple. I needed to, I needed to get some some prototypes made. And my usual supplier at the time was, they just weren't responding to my emails. For whatever reason, my my singular contact was just non responsive. And I needed to send a board order out for quick turn that day, like literally by 5pm, that day, that order needed to go out. So I had to go and find someone else. And it was I honestly, really two choices, which are the same two choices that I imagine most anybody else in my position has, which is one, go to Google and search for circuit board manufacturer or PCB fab and then go down the list of the, you know, the 45 ads that immediately show up and then who paid the most amount of money to be shown up there. Exactly. Or you go to PCB shopper, which is basically the same thing except in some hierarchical format that I guess it's sorted by cost. There's what it boiled down to was, I was presented with choice, but I wasn't presented with any way of making a decision about that choice. And my decision was based not on cost because I had essentially unlimited budget for that project. It was based on delivery. Can you deliver this in my time for him, and everybody promised that they could. So if cost is not enough, is not an issue. And delivery is not an issue. You The only thing that's left is quality? And how do I, how can I pick quality, when the only thing that's available to me, are these kind of questionable five star reviews that everybody seems to get, that tend to be in somewhat broken English, you can't really make an informed decision. And I was really frustrated by that. And this was specifically on PCB shopper that I couldn't pick any one of them because I, I felt that they weren't good, because they all presented the same information. And the only thing that was different was price. And I kept looking for somebody that had published a picture of the board that they had made, not because Oh, hey, this is the board that I made. More like, this is the board that PCB way made for me. And this is the quality that I got back and no one was presenting that there. I think EDA online, or maybe e times I think they wrote some articles a while back where they had done some circuit board reviews, but they wasn't of sufficient detail, to really, to really get me to to a spot where I thought I was getting the information that I wanted in order to make a decision. And as with many things, I, in my frustration, I'm sitting there, you know, kind of screaming in my head, somebody should do something about this, like, well, I should do something about this. So I started to sit there and think about what what I need to do in order to actually deliver this in a way that is useful and trustworthy. Because I think those winds up being the two key elements to this useful is, and like, am I covering the information that that I would want to see and trustworthy in the sense that if I write a bad review for a manufacturer, can, am I sure that that manufacturer is going to come back? get pissed off reading that review? But go he's he's right. I mean, I don't like that it's out there. But he's, he's right. So those were the two, those were the two, the two things that I really wanted to kill on for this site, make sure that the content was quality, and that it was detailed. And that I was basically that I was showing my work that there wasn't any question about how I was going to be evaluating the boards or the manufacturers?
What criteria do you look at then? Like, it's kind of like, you know, just saying it looks good is kind of subjective.
Absolutely. And I have to be honest, that in some aspects, subjectivity cannot be avoided. So I try to cover things from not just the board side, but on the whole interaction with a particular manufacturer. So what is their customer service? Like? What is their ordering process? Like? What's the fulfillment like? And I mean, that stuff? My opinion of how easy a website is to use is just that it's an opinion, Parker might have a different one, Steven might have a different one. So those I absolutely subjective. Subjective, I think, from the standpoint of I am a web user in the 21st century. So I think we all kind of assume a certain amount of ease of use. But, you know, can I just go to the site? And is, are the design rules easy to find? Do I have to jump through a bunch of hoops to, to get to the point where I can actually order something? Do I get a JSON error when I when I try to send you my money? Like those sorts of things are, in fact, subjective. So yeah, from the on the business side, I fully admit, there's very little I can do to avoid that subjectivity other than saying that, chances are my opinion is probably in line with where most people would be. But I also thought as I was putting these design are these review criteria together, is that there are probably a lot of people out there that would look at a review and go, You know what, I honestly don't care how friendly customer service is, if I get good boards. Like, for me, it's very important that you give me an invoice that I can expense. And if you give me an invoice that's completely in Chinese. I'm gonna have a hard time submitting that to any accounting department on Earth, where at least well in the United States, and actually getting my money back from that particular client. Yeah, it'd be fine in China. It would be amazing in China and fortunately, I don't have any consulting clients in China. The clients that I have here in Los Angeles turned out that they actually want something that they can read themselves to verify that they're You know, reimbursing me for what it is that I'm telling them, they're reimbursing me. So for me, it's very important that an invoice be readily available and readable. And the sort of thing that I could submit with an expense report, that might not be important to anybody else on Earth. But I'm the one writing the reviews. So when it gets to the board's that's where I try to remove all the subjectivity as much as I possibly can. And I do have on the site under the Resources area, all of my review criteria are listed under there. So the way something gets scored should be completely listed out based on the specific tolerances that I'm measuring to, and the end the measurements that I'm expecting. And in some cases, the basically all of these criteria are coming from two spots. One, every manufacturer is publishing the tolerances that they are willing to advertise to, all of that gets published, or I've got the IPC, I can go to the to the IPC and read what their documentation is, or I can choose what the manufacturer is publishing. And for the most part, I'm selecting the manufacturer, because that seems to be the best way to do it. And I am, I am under no circumstances, expecting perfect boards, because I am ordering deliberately low quantities of boards, from low cost manufacturers in the way that I suspect a lot of people that that are coming to this from where I was maybe, you know, five years ago, 10 years ago, that they're coming into this and and I, I want to allow somebody to, to eliminate the idea of is it Did I screw this up? Or did my manufacturer screw this up? Because I've had instances where you know where I've been before I started manufacturing rooms where or manufacturing reports where I got a board back and soldered it up and tested my circuit and couldn't figure out why it was working. And it wasn't until I put it underneath an inspection lamp that I realized that at some point, some drop of some fluid somewhere landed on top of a trace. And and it forced that trace to get etched through when that should have been a continuous signal. And that's that is a pure manufacturing fault. It obviously means that they didn't eat us the board. And you know, what's what's my option then is to you know, just to wait for these things to come back after they have been remanufactured provided that the manufacturer is actually going to do that for me.
And it's not just that it's because I've gone through this before is you've already put all the parts on the board, and the expensive part of manufacturing, which is placing the parts and reflow and all that stuff home and the time. And then like yeah, like a $10 PCB at that point is chump change. And right you ask the PCB manufacturer to give you new ones and the lecture will give you new ones and you're like, Okay, now I got to spend another 120 bucks on components now, right?
Yeah, exactly where I can try and rework all of that stuff off of there, which is going to cost me again, time and money. And it's that, as I have discovered over the years, particularly as I've moved more into working with electronics on a daily basis, and it being the, the way I pay the rent, is that I will pay $100 for a board that works the first time from the manufacturer, rather than 10 cents for boards that I have to get respond because there was a problem in manufacturing, because what I'm losing is time. And time is the one thing that I cannot generate it from anywhere. Once I've lost it, it's gone, as opposed to money where I can just work harder and make a little bit more money and, and try to get things to line up as best as I possibly can so that I'm working with the best materials that I possibly can. And so, you know, the when I, when I decided that I was going to actually pursue putting this site together. I honestly had no idea what I was going to get. My biggest fear was that I was going to send out my I did a an initial run of of 10 boards with five different manufacturers figured that would be a good test scenario for seeing what the differences were. And I strongly suspected that I was going to get 50 boards back and there wasn't going to be a single thing different across any of them. And my grand plan for reviewing these boards was going to be a single page that said just go with doesn't miss one. Yeah, exactly. This is in fact a commodity. Well, that would be the that's what
you expect, though. I mean, obviously reality is not going to be that but
as it turns out reality isn't what Make that at all. And I was, I was scared that all the boards were basically going to come back exactly the same. And I am shocked to discover how different they actually are from one manufacturer to the next.
And manufacturing reports.com is sort of your detail of all of the differences between them.
Oh, yeah. And somebody, I think it's a blog in Italy, for whatever reason, somehow, in the first two or three days of the site being live, wrote that, in terms of, I think it's like a chi CAD circuit board design page. And somewhere down towards the bottom, he says, you know, if you're looking for for a PCB manufacturer, here are a couple that I recommend, or you can go to this new site manufacturing reports.com, which goes into, quote, painful detail on the quality of the boards that are reported, I thought, is that a, is that a good thing? Like, like painful, excruciating detail, but like, well, it's, it's a lot easier to like, scroll through the information that you don't care about, rather than sitting there wishing that something had been published instead?
Well, and okay, so there's, there's images, yeah, every, every board all the way down to the packaging. So you can see how it's going to come from, you know, your detailing to everyone in the world. If you order from XYZ guy, here's what you should expect is what you're gonna get. And and the whole the thing I don't think we've really mentioned yet, you created what you call a test coupon, which is just basically a universal board, right? That that you sent to all of them to just be a kind of a fair test. And none of them knew that this was a test versus the others. Right?
That is that is absolutely true, I struggled that I kind of struggled with this, because I, I hate the idea of just like creating e waste, there just doesn't seem to be any point in creating a board. That doesn't work. But ultimately, in the end, I couldn't figure out some design that I could send out. That would have a practical purpose, other than for me to just measure the crap out of it. But it had to be standardized. I wanted it to be something that that wasn't a torture test, a torture test isn't fair a torture test is how well can you live up to the to the best quality manufacturing services that exist on Earth. And that's not what I want. What I want is, are you giving me what you say you're giving me that's the key. And the board is designed around making those sorts of things as easy for me to measure as possible. So you know, the, the outside dimensions of the board are easy for me to measure with calipers. The copper thickness is designed so that it's easy, at least in version two of the board, easy for me to measure with micrometers, I've got an area on there where it's easy for me to get in and measure the width of traces. And then also to have the variety of features that I know people need you, you need to know that if I, if I if this board is going to be 50 millimeters, and it's going into an enclosure, like that board needs to be 50 millimeters plus or minus you know, a certain tolerance. The same thing with with mounting holes like those if I'm expecting to put an M five screw through there, that screw hole better be you know, within point oh one or point oh two millimeters of 5.3 millimeters. Otherwise, my physical screw is not going to physically get through the physical hole that these guys supposedly milled on the board. And, and as I've discovered, some people are really good at some stuff. And some people really are good at other stuff. And turns out some people are absolute crap at doing other things. Which shocked me, like, you know, again, I kind of expect that if I'm going to spend money on something even if you are selling it to me at 90 cents a pop that I'm gonna get a certain level of quality in return for it.
So do these manual please because you've done five so far right?
I'm up to I have six published I've got a seventh actually on the test bench right now.
All right, so have have a PC manufacturer come back and been like, this looks like a test.
Nobody has responded with a like this board doesn't make any sense. You have disconnected nets. You've got this weird sort of accordion strip on here that has you know, traces that step down in width from from 20 mil down to six mil but they're not connected to anything like nobody and and to be fair, I did kind of design the board so that it would almost look useful. Like there's a QFN on there. There's a qf P there are traces that lead off the qof P two v is they are there so that I can see how well you actually hit the center of that via and, and how what's the quality of the Q SFP that you're actually producing. But like there's a whole line of passives. On the left hand side, they go from 1206 down to, oh, 201 they're not connected anything. It's just a bunch of like bare pads that are just sitting on the board.
It's pretty obvious that it's a test PCB.
If you look at it, yes.
Yeah. Someone like us looks at it like yeah, it looks like a test PCB. But especially like the USB connector, it's got traces that go nowhere, they literally
go one centimeter and, and disappear into the solder mask. Yeah. So yeah, there's there's an interesting point, actually, today wireless
So nobody has come back and said, Hey, this appears to be a test coupon. I did get an email back from all PCB asking me whether some of these features that made no sense were actually necessary. Like, we can't figure out any reason why you have a solder mask keep out area in this particular location. Do you need that? Like, or is that a screw up inside your Gerber's like that? That is absolutely vital for testing. Thank you very much for asking. And then they fill that area with solder mask somehow.
For that, correspondence, I guess, you got that from one of the people he tested, do you count that as a positive or negative,
that was an absolute positive. And I was, I was so looking forward to testing their boards thinking that maybe I've actually found the unicorn manufacturer here that I couldn't bring myself to do it for a couple days, I didn't want to dispel my fantasy, that these might actually be really quality boards. And then I opened up the packaging, and I looked at the boards, and I didn't even have to put them under magnification to go Ah, come on. That's the thing. Like, if if you look at the board, you know, it's a test coupon. But in order to know it's a test coupon, you have to look at the board. And whoever is running their line, there's no way in hell, they actually looked at these boards, that solder mask is so scummy, in that keepout area, I mean, it I don't understand how I actually do not understand how it can be manufactured in that way. I'm assuming they're using the liquid Photo Imaging, they're not using dry masks and aren't using dry film. They expose it, this is a big chunk of, of black, right, that is not going to get exposed. They run it under the UV light, that ink cures, they lift up that mask and then they they run it through whatever their solvent is to clean the liquid stuff off. And it didn't work. Maybe their solvent was old, maybe it was contaminated. Someone sneezed in it exactly like there was so much that in on one board, there was a complete film of solder mask leftover in the keepout area that prevented me from actually measuring the traces the width of the traces in that area. And in the others, it looked like somebody had blown their nose, like right across that that particular part of the board. And it the thing that struck me is that you don't even have to look hard. All you have to do is look to know we screwed these boards up this board it there's just no way you can look at that board and go this is an acceptable thing to send to my customer. So all PCB did ask about the features on the board. Nobody has asked Is this a test coupon is are you evaluating our manufacturing services or anything like that. And in one instance, this just happened this morning, a manufacturer has written back to me to say I have seen this review that you have written about me. And I was a little scared to open that email because I'm assuming that at some point, I'm going to start getting more of these, but it was the first one that I'd received. And I was getting ready. I was preparing myself for a bit of an ass joing. And it turned out that they said we read your review. Thank you for being objective in the way that you evaluated our boards. I have gone to the factory floor and I have talked with the people about our via drills and about I think it was maybe something about the solder mask offset. We'll be looking at that. And we just wanted to let you know basically that we saw the review and we're going to be talking about this internally to see if we can improve these processes. And like hey, right on, like if, if this results in making positive change at a manufacturers let people get better quality. I really couldn't ask for more than that. I mean, I could ask for advertisers but yeah Another point. But as far as, as far as you know, having whatever little effect i can positively have on the industry. Hey, that's a that's an excellent first step.
That's really cool. And I'm curious, would you ever be willing to redo boards with, you know, that company and test it out? If they got back to you and said, we've adjusted our process? Can we run boards from you for you again? I mean, that might be a little bit, they might skew things because they might put more effort into it. But would you be willing to do that?
You're not asking any questions that I haven't asked myself. And when I when I was thinking about how to evaluate customer service, I thought if if a manufacturer if I'm if I am Dan Hinch of Reinbold heavy not Dan Hinch of manufacturing reports, and somebody sends me crap boards, I am obviously going to respond to them that hey, you sent me crap boards, you better send me better quality stuff. And then on that second turn, I'm kind of expecting somebody to maybe put a little bit more of a microscope on the process than they did on the first round. And I'm not giving them that opportunity. This is the I gave you money. And I gave you my specs. And first shot. Tell me like show me what you're given me. And then I thought, you know, like when I go look at a camera review. It's not like there's, you know, the reviewer of the latest Canon DSLR went, I really hate the sensor in this thing. And then went back to Canon and spent six months on the phone with their sensor designer until they got the camera they were expecting. Now they went out they bought the camera. This is the camera that I bought that most people should expect. And here's what you get. And, look, I will absolutely go back to these manufacturers. But I'll go back to these manufacturers after everybody else has had their turn. So yeah, I am I am very grateful that this one part, I don't know why I'm being anonymous about it. It was PCBWay PCBWay came back and said, we saw the review. We're going to go investigate these processes on the floor. I will be absolutely happy to go back and do another review with them. But I've got others that I gotta get to first. You know,
so if you're looking to buy boards, or boards right now go to PC way. flying on an airline, that's after they've had a wreck or crash? Because that's like the most safe airline now.
That's exactly what everybody was thinking of Ethiopian Airlines, right? Oh, yeah. Lion Air. Yeah. I'm kind of hoping you guys are laughing because it froze right there. Yeah, we were, like, completely landed with just absolutely the wrong. Like, hold the pin on the grenade and
Wi Fi knows how they've meetly dropped for the most
comedic effect. Yes.
So I got a question. Why hassle?
Because everybody offers hassle. Okay, so if Well, I take that back. Hassle is the default, I should put that hassle is the default at most manufacturers. There are very few where you are forced to go with with enough egg, or with the whatever the environmentally friendly acronym is, I don't remember what that is.
Oh, S P or
something like that. Yeah, exactly. So I wanted to, I wanted to do something that I thought would be most available across across them. Exactly. And also, it also winds up being invariably the least expensive. So the people that are out there trying to get budget boards, when they discover that, you know, I can pick, I can pick Hassel as my surface finish. And I can get these boards made in 24 hours, and delivered in five days. Or I can choose enig. And I'm guaranteed the super flat planar surfaces, but it's going to add four days to the manufacturing time. And it's going to add, you know, $30 to my cost. As soon as they see that they're gonna go back to selecting hassle. So while I personally probably wouldn't choose hassle for any prototypes, or certainly not for production boards that I'm manufacturing I'm trying to get into the head of the people that are that are at least at this at this early stage of manufacturing reports. What the What I'm trying to be as inline with the selections that most people would be doing and I imagine most people are going to these incredibly detailed ordering pages and like ah, I'm just gonna pick the defaults and hope that I get what I expect. But yeah, I mean, but then again, I have gotten some hassle that's a really weird phrase. I have gotten some some hassle surface finishes by ACC that looked like they are plated the the ones that I got from all PCB were gorgeous. Like you would thought that they had plated that stuff in silver. It looks so beautiful. It didn't wick through the vias at all. There was very little like surface ripple on the surface. It was just smooth, smooth, smooth, smooth, smooth. And then I had these boards from JLC PCB that look like they just basically used a fire extinguisher to apply the solder to the board.
I've seen boards that I've gotten boards that look like they've been dipped in liquid tin, like a wave solder machine. Yeah, yeah.
And I went. So one of the features of the site is a glossary of manufacturing terms. And I'm trying to like get into the deep detail on it on a human understandable way for what each of these features are. Like, if you go to the Wikipedia page for hassle. I would imagine that it's probably like the detail on the molecular structure of solder. And really, who cares? What I want to know is what in the hell is hot air solder leveling and and what does it look like? What's the difference between it that and evening and I found a video of of a of a hassle application somewhere in China that looks medieval, like this, this poor PCB gets put into this rusted cast iron machine. That just like plops it down into what I imagined is probably about 100 litres of liquid solder and then like drags it back up through a couple of air blades. And then it it clunks to a stop sits there in shivers and you can see solder dripping off the bottom of this board. And then the technician reaches over with these thick gloves on unclip it from that throws it onto the assembly line grabs the next one. And the process repeats. Like it's I'm kind of assuming that some of the boards that I'm getting back are going through that process versus something that looks like it was actually post industrial revolution. So yeah,
we've talked about before, there's a video out there of a old school transistor clock and clock radio being built. And it's a manufacturing line from like, I think like probably the early 60s. But they have a their wave machine is it looks like a Ferris wheel. And then they put the lazy donut the boards into flux and put them on this carousel and the carousel goes around. And then it just slaps the water the not water. The liquid tin. Oh, I guess at that point would be led the LED and it goes this huge plume of the flux vapors going off. And then as it comes up, someone just grabbed it off them off the Ferris wheel with no gloves on puts it on the conveyor line. Oh yeah, no
gloves. And I guarantee you they didn't have any like respirators on or anything like that. And of course not. And you know, it's that like ear wax are a flux all over that stuff. So the smoke from that stuff is awful.
And that's why you have to dunk the whole board in that lead barn at all. Yeah.
Well, you know, they were probably smoking a cigarette while they did it also anyway, so
Oh, well, yeah, of course. Well,
it's probably that was the that was the respirator they're
filtered through? Exactly. they hand you a pack out your way into the factory floor. Right, you will be healthy provided that you actually smoke that full pack before your shift is over OSHA brand
So I've got a question for you that I'm this is one of the one of the first things that came to mind and I'm really interested to hear what the results are is did any of the board houses make changes to your boards, like so take for for instance, if you provide a board to a house and you have silkscreen that runs on top of a pad, most of the time they will get they will remove the silkscreen from a copper area. Sometimes they don't though, but technically they are changing your boards or modifying them somehow. Did you notice them doing any of that?
So the flip answer that is absolutely because any Chinese manufacturer that you're going to work with always puts their damn manufacturer's Mark somewhere on your board even though you did not ask them to so and I think it is PCB way where they actually have an option to turn that off. And they charge $4 for it
look at the your board now and figure out what you want it is
you know I take that back. I don't remember the what was oh it was electro electro has an option for dollars to turn off. Having them actually print their manufacturers mark on the board. PCB way does not have that option. But what they did do is they printed it inside the qf P footprint so that When you solder your part down, it's not visible. I'm like, that is the like, we can't stop doing this. But we'll at least be conscientious about how we're going to apply this. So yeah. So invariably, when you're dealing with offshore manufacturing, most likely, you're going to get some kind of artwork on your silk layer, either the front or the back, that you did not request. Again, this wasn't a, I'm not trying to torture the manufacturers, and I'm not trying to give them gotchas. I thought about that about like, you know, just draw a big line across a whole bunch of footprints and see whether or not they go ahead and try and manufacture that. That's not a realistic test. Because who would do that? I mean, yeah, I've I've made that mistake, I'm sure at some point where I didn't look at the Gerber enough. And at some point, I rotated the reference designator right on top of another Pat. So when it comes to like the silk layer, no, and I also didn't deliberately try to do that to them. But I do see changes to the solder mask relief, that whatever that aperture sizes around the around the pads, they will change that to match whatever their tolerance is within their manufacturing process. And sometimes they'll also change the the copper port dimensions to match whatever their their solder mask tolerances are. And then sometimes they don't. So the manufacturer that, that I just got boards from, have the ground, pour on the backside, just completely exposed from underneath the solder mask. And I thought, well, that's an obvious screw up. And then I went back and I looked at my Gerber's. And on in version two of this board, there, I did not provide any relief, like my keep out layer is, or my keep out geometry is exactly the same as the ground poor geometry. So I can't thank them for that. Like they're, they have tolerances that they manufacture too. And if I'm expecting them to keep perfect tolerances, but I did not tell them that this must be perfect tolerances, then that's my screw up, not theirs, within a certain range, like they also publish that tolerance that I get four mil of play, or the manufacturer gets four mil of play on that solder mask. So I'm going to grab my machinists comparator with a reticle on there, that's got a four mil line. And I'm going to put that over there and measure whether or not you actually adhere to those tolerances that you said you would. But again, I don't deliberately go out to try to build something where they have to catch it. Again, on on version two of the board, I did screw up and put the solder mask, not on a pad, but outside, but I put this the silkscreen in a keep out area where they would have had to print it actually on the substrate not with any copper there, but like directly on the FR four. And, and one of the manufacturers did was all PCBs. Like, do you want us to really do this? Like, no, that's my screw up, go ahead. You can just ignore that solder mask just eliminate it from the from the Gerber, that's fine. The other manufacturer did not ask me. They just wound up not printing it. So yeah, they went out of their way to not match my design without telling me. But again, like I don't, I don't like that idea of me holding somebody's feet to the fire. Because they didn't fix my screw up. Right? Like I should be giving them my best possible design. And they should be giving me their best quality services based on the design that I gave them. And if it's a screw up on my part, it's a screw up on my part. But if it looks like you, like wipe the toilet paper of silkscreen ink across my board, that is your screw up Shenzen to you.
My, this runs through a story of back and kind of like the beginning of macro fab like so it was like four years ago with Steven was working there. And at the time, our PC manufacturer was putting their logo onto our boards and the UL marking and we're like, and that's kind of lame. So we I we actually called them up and we're like, Hey, can you just put like their macro fab logo and so your logo and they're like, yeah, we'll do whatever. worst mistake ever. Like the first panel that rolls out that gets shipped to customers. phones are ringing like why is macro fab on our boards. Steve and I thought this was like a brilliant idea. Like for marketing. We're like, Man, this is me amazing. Like people take pictures of the boards. It's gonna have macro grab their worst mistake ever. Yeah. I had to reprint like, what, like 20 or 30 jobs worth of panels.
You know, it's funny, so that that company that actually gives you the option to pay to have the manufacturers Mark removed, you do not have to pay to have them put the UL listing logo on your board. Which to me just kind of sounds a little hinky. Like, why you're just offering to just give me certification logos on my board, just, you know, I mean, nothing, nothing preventing me from downloading that and converting the PNG and vector graphics to put on my board myself, I suppose. But it just seems an odd. Like what you just like happened to have this vector graphics sitting around. So we just put it onto the ordering options. Like, you know, it's it's there, like, you know, do I want to read solder mask or green solder mask? And do I want to just assume that my board is UL listed? Well just check that box, then we'll we'll we'll throw it on there. Wherever we can find space.
You know, I think it would be fun to talk to someone about you will requirements on the raw PCB itself, because I think that comes down to not the assembly. Like if you have the UL mark on the actual PCB. I don't believe that that applies, like across the board to your whole assembly. It's it's saying that the board house that you purchase it from follows a UL process? Well, and you know, or at least they just stamp it on there and say that they follow that
process. Yeah, exactly. So it's actually speaking of industrial certifications, I've and I'd be interested to hear what your guys's opinion is of this. It turns out that there is a organization here in Los Angeles that that does IPC training and IPC certifications. And I've been thinking about going for the IPC, IPC 610 C, which is the assembly inspector or the assembly Specialist certification. They offer I think it's like a three day course for the certification. And then it's a one or two day lab, where I'm assuming they give you a bunch of boards that you have to sit there and analyze and determine whether or not it matches qualification or not. Is that are you guys familiar with that? Do you think that sounds useful?
It's incredibly useful.
It's incredible. Yeah. The majority of like our QC team, and people on our line, our IPC 610 C.
Right, and you walk away with the, with the book, or at least a digital copy of the book. That is basically a the more in depth version of your website.
Yeah, exactly. And on my like, on my original review criteria page, up at the top, it's, it's it did used to say in big bold letters, I am not IPC qualified. Like, I am not trying to say that this is a class two class three board, I'm not trying to say leeches meets any kind of particular assembly criteria. These are just the criteria that I came up with. And then honestly, I thought, like, Why do I have to devalue myself? I'm not gonna go out there and say, I mean, I could go out there and say that I am not IPC certified. I could also say, I'm not ISO 9001 certified. Nobody cares.
I would say actually, that would, if you got that done, that would lead more credibility, because like, you can say, Hey, I am XYZ. And this is why my opinion matters. Now
I have a piece of paper that says I matter.
Basically, if I if I had just gone for that IPC certification, beforehand, I could just like throw all the detail out the window, because you don't need it anymore. Now, you can just rely on the piece of paper that I've got hanging on the wall. And I could just have a picture of the piece of paper with like a thumbs up or a thumbs down over the name of the manufacturer, I just have to trust right now it's not on me. Now it's on IPC. If IPC says, I've got the authority to say this, then obviously I do.
So, um, I was actually going through before the podcast through all the criteria and all the listings and stuff. And one thing I would like to see so this is kind of like some feedback from me is, I like to see like a resistance to like reworking the board. Sure, especially if you're doing prototypes, you're definitely going to be unsought during soldering back on parts. And I know some of the cheaper boards are running less expensive substrates and you get bubbling and and d lashari and stuff like that, whereas, like, for example, like OSHPark runs like TG 170 boards, which are more resistant to multiple reflow processes.
So on my the The covering the solderability and and rework capability or rework. capability is not the right word, but the the survivability of the work of these PCBs. It has been something that I've evolved over time. And what I've boiled it down to is, how, how easy is it to solder, a USB micro connector or a USB mini USB micro connector to this PCB? And how difficult is it to hand solder a qf P to this board. I thought about doing QFN. And then I thought, well, if I'm going to do QFN, now I've got to do solder. Now I've got to do paste stencils. And now I'm starting to introduce a whole bunch of variables that don't really have much to do with the actual PCB itself. So I should probably stick to things that most people can hand solder. So what the process that I've hit on now is to is to solder the qf p and then rework that chip twice, and see what the effect is on the board. And in for this, I don't give it a score. Because I, I don't know how to give the the rework a subjective score, like, invariably, the solder mask starts to erode. But like what is a what is a five point erosion versus an eight point erosion. And if I'm going to talk about the quality of the board, rather than the subjective stuff I was talking about earlier, I shouldn't score that if I don't have some way of backing up that score. But I know this is important. So I want to at least publish it without necessarily scoring it. So you can see what happened when Dan in his crap soldering and rework skills, actually sat there and started to, to heat up and cool down this board. So I do that with the Q FP. And I do that with a QFN. And, and invariably, the solder mask starts to erode. I haven't seen any bubbling, I haven't seen any pads lift. I haven't seen. I've seen a couple boards start to smoke a little bit. Although I don't know whether or not that is necessarily some evaporation coming off of the solder mask, or whether that's just some flux that happened to be still underneath the chip that's starting to boil off suddenly. But they've they've, they've all been pretty resilient to rework. The solder mask, not necessarily so much. But this kind of leads me into something that that delayed the launch of the site for a significant length of time, because I became obsessed with trying to characterize how much force it takes to rip a trace off the board. And that was the original point of the micro USB connector was that I was going to solder that thing down. And then I was going to stick a cable into it. And then I was going to measure how much force it took to actually pry that SOB off of the PCB. And I spent at least three weeks and I'm talking not like three weeks in an office like three weeks as a self employed person, not getting paid to do any of this stuff, like 1012 hours a day for three weeks straight, no time off, trying to come up with some kind of jig, where I could apply consistent pressure, and then measure the breakaway force. And I got close. But I never got to the point where it was repeatable. So I eventually just had to pull the pin on it and just abandon it. But I still actually got the jig sitting on the floor. And it's this like this, this open be monstrosity, seesaw with a with a with like this, this medieval screw that comes down from the top and sits there slowly like trying to rip the USB off the board. I got close. It's it's in my plans of actually like figuring out
if there is a difference between the manufacturers, because based on the results that I did get that I didn't feel comfortable publishing is that there is in fact a difference. I just can't I just can't say conclusively, and I can't say it with authority. And I never got to the point where I was comfortable with like yeah, this happened like four times five times in a row and like it is the exact same, you know, amount of grams or kilograms of force required to pull these to pull This connector free. But this, this idea of how How, how strong is this? Or how susceptible is this board to breaking if I actually tried to de solder stuff off of there, or if I put my chip down wrong, or if I wind up smoking that chip because I screwed up some other part of the board, I thought that was very important. And I wanted to at least present it. Even if I couldn't really find a way to characterize it
sounds like IPC would actually probably give you a little bit more authority on saying those things,
probably. And mind you, I, I, I am I am. Again, I am not an IPC certified Tech, I don't have their their documents. The best I've been able to find have been abstracts and summaries, and, you know, a couple of people that managed to publish, you know, some PDF that's probably close to like 10 years old at this point. Which is why I'm I am not adhering, or at least I'm not saying that I'm adhering to IPC standards, because I don't have the authority to say that, I will say that my opinions are informed by the IPC standards. Yeah, exactly, to the best that I can. But for the most part, I'm trying to rely on what the manufacturers are at least telling me that they're going to do and if you tell me that you're going to do it, I'm gonna hold you to it.
So what is the worst piece so this doesn't have to deal with manufacturing parts, what's your What is the worst PCB you've ever seen? And it could be like, when you open up like that forklift, as well,
I tell you, the, the worst. The worst PCB I've ever seen, worked absolutely beautifully, and it was beautifully manufactured. And it was, it was beautifully assembled. But the design of this thing, such a nightmare, of, of like precision bent angles on on pin headers, so that so that your sensor could be at just exactly the right angle to receive a reflected infrared signal, like it was broadcasting infrared, and that would reflect off something else. But the infrared receiver had to be adjust exactly the right angle. And nobody happened to manufacture, like pin headers that are bent to 23.34 degrees. Like so. So I had to sit there and like bend the pin headers, these specific angles. And I'll man that thing was was an absolute nightmare, but that had absolutely nothing to do with the manufacturers. I have opened up some seriously crusty, like 1980s era remote controls. And you know, these, these, these single board design or these these single layer designs that have these, I mean, to start with the trace design on these things, when they were still hand drawn in ink are gorgeous, like they are super organic looking, but the manufacturing process, man, you can still smell the flux. And those things are crusty. Like, like, I mean, there is a there is a physical crust on these boards that you could pick off with your fingernails. I remember my first job in it was working for the General Services Administration. And I had to field strip 386 PCs at at the, at the region nine headquarters in San Francisco. And you know, their token ring card would go dead. And you'd have to grab one off the shelf and run out and swap it out on the floor. And I would come back with these, these, these needle marks in the edges of my fingers from the solder points on the back of those boards. Because the I mean those those things were all through all components. And I don't know what kind of wave soldering they were doing or how they were clipping those leads. But it was a field of hypodermic needle points on the back of those boards that you had to pull out of those PC cases. And I tell you what, you never let go with a thing because it was literally embedded in your flesh. Oh, as you were trying to yank it out of there. I did get a board back from OSH Park. And this was this was years ago now. And and they had something had happened in The manufacturing process that when I went to solder my prototypes, the solder just balled up for would not adhere to the pads. And I thought maybe there was a problem with my solder or the iron or the flux or something, and I ordered, I don't know how much difference well, at the time, you can really buy that stuff from Amazon. So I was like, going out to the local supply shop. And, and and, and, you know, just sort of replacing my supply flocks replacing getting a new roll of castor solder. And it turned out something had happened in the manufacturing process where there was this weird, thin coating of something on that board that refused to allow solder to adhere to the pads. And it was weird, like you would sit there and you would solder and you would solder and you would solder. And if the ball didn't just wind up sticking to the tip of your soldering iron, it would stay on the board, and it would just form a sphere and you could flick it off with your fingernail. It was the strangest thing.
Someone decided to pre conformal coat your boards.
Exactly. But, you know, I sent an email to support at OSH park.com and said, I don't understand at all what's going on with this? And I said, No problem, we'll send you new boards right away. And I tell you, so at least as of the recording of this podcast, OSH Park does have the number one score on my site, turns out their boards are really good. Their customer service is so as part of the review process. I have some I find some reason of contacting customer support. Invariably, I'm asking them for a copy of the invoice or what's my ship date? Or did you get my order? I find some reason to contact them specifically, so I can find out how long it takes them to respond. Is it easy to understand the communication? I sent an email to OSH park at 6:45pm pacific time on a Friday night. And I got a response within 10 minutes. And I thought yeah, no one's gonna beat that. And no one OSH Park really honestly cares enough to respond to those emails. And most people don't. So, so yeah, they got they actually got an extra point for that, again, because it is entirely subjective. And I do the scoring. They want of getting an extra point for just being that diligent as opposed to somebody that would just you know, if you, if I send you an email at 615 on a Friday, I'm not expecting you to respond. I'm waiting until Monday to get that email. So if you actually respond on a Friday night when honestly, I expect you to be drinking a beer and eating pizza. Hey, you know, thank you very much. That's, that's, that's absolutely going above and beyond, as you know, all PCB did with with with asking me specific questions about to design files that I
sent, and then doing the exact opposite? And then
actually, yeah, not adhering to the thing that they should deliberately be about.
So what's next, for manufacturing ports?
Well, my, my, my, my goal, or my original goal, and I think this is a good thing to stick to is to go through the list of 20 manufacturers that I've identified and get boards from all of them and publish reviews based on the criteria that I have set. I've I I think that's a solid course of action. But what I would obviously like to do is to move into other areas that I think people are interested in, in of the two that I thought initially. That would be important. One is flex PCB manufacturing, and the other one is PCB assembly. A few people have asked me about, are you going to go into evaluating flex PCBs? Everybody has asked me about assembly without question everybody wants to know about, like, which assembler Do I go to? Because all of these low cost PCB manufacturers, all the people who go to for low cost bare boards are also advertising assembly services. So that is absolutely something that I want to go into. I think it is in my personal best interests to make sure that I've got that IPC certification under my belt before I actually go and do that. Because my suspicion is that I need to know more about that, about that evaluation process. Then I can just figure out for myself, which is kind of like it's kind of easy to characterize these boards, you know, like a bare board. All you really need is to actually care enough to measure these things, using tools that you can buy freely available on the marketplace. For not much cost, you just have to care enough to do that. Right? When it gets to doing the assembly, or evaluating assembly, I think I'm going to need to have a little bit more authority on my side in order to do that.
Yeah, the one thing I was looking at was when you're measuring the trace widths and how off, they always
are, oh, my God,
it's like, man, if you're calculating like, right on the threshold of like, you know, like a 10, to see temperature gradient on the floor power supply, right? You might want to not use the cheapest person,
I'm telling you. Like the, I think the first two or three manufacturers, I looked at all of their traces were within one mil. And I was like, Yeah, I mean, that's, that's not just down to the quality of their manufacturing, that's down to the tolerances of my measuring equipment at that point, because the calipers that I'm using only measure down to five tenths of a 1,000th of an inch. Now I've I've changed Well, I've, I've augmented that capability. Now with this machine, it's comparator, where I can do an actual direct visual comparison to a standard. And I can physically see whether or not your trace is actually six mil, or five mil, or four mil, rather than sitting there trying to rely on whether or not I can measure I can directly measure correctly with my caliper. So the my new measuring process is to measure it with the calipers. And if you come in significantly, well, if you come in under tolerance, that I am going to go and do the visual comparison then. But yeah, but these are the things that I just assumed were correct. I assumed that if I specify a trace width, I'm gonna get that trace width. And, and this is the thing that absolutely blows me away, I assumed that if I ordered a specific thickness of copper, I was going to actually get that thickness of copper. And what I'm discovering is that in some cases, you order one ounce copper, and what you're getting is something more along the lines of half ounce copper. That's bad. Yeah, that's not cool. Man. We
what's interesting with our PCB manufacturers that we use at macro lab is usually you actually get a view order one out, so you actually use to get slightly thicker copper.
And I've been waiting for that, in some cases. And because I'm also not, all of these things are coated with the surface finish, right. So like, especially since all of these have got hassle on them, the only one that had Enoch was OSH Park, like if you put hassle on there, I don't know exactly how much thickness you're adding there. But it's it's more than a little bit. And your pads are still coming in undersized, like that thickness is still remarkably thin, you should be with one ounce copper, you are at, I think 1.4 Mil Standard. And these guys are coming in, in some cases that half of that, like that's, that's just not cool. If you're giving me copper that thin. And then you're also giving me really, really thin traces on top of it. Like, I can't, I can't expect to get any kind of decent signaling through that. And I certainly can't expect any kind of current to flow through there. My boards just gonna vaporize on the spot.
That really cool test to add is a like you couldn't have a current past test and measured temperature rise.
Yeah. So before we were talking about what is next, yeah, I would really, really love to send a for destruction only coupon to these manufacturers, where I've got, you know, traces that I can subject to basically a short circuit and see how much how much current they're capable of carrying, go through the go through the calculations and figure out if, you know, if, if they're, if I'm specifying this trace width, and they're saying I'm getting this thickness of copper, how much current does it take to actually burn that trace off the board? Again, that like how much force does it take to lift a trace off the board? I think there's some interesting destructive tests that could be done to characterize some of the stuff that they're manufacturing. But again, all of that stuff needs some significant testing, because I don't think there are necessarily standards for that, at least not that I'm aware.
Yeah, when we call any either. You also might realize that all the trace with calculators are bunch of bullshit on what on?
Yeah, you know, and then that's the like, what, you know, when the when the trace actually burns off the board, is it the fault of the board? Or is it the fault of the calculator that I used? Exactly. All right. ate like those. That's why That's why I've been very careful about about establishing the review criteria for this stuff because I want to make honestly, I don't have an axe to grind with any of these manufacturers, I really desperately want to be able to shout from the rooftops, like give your money to these people, this is who you should work for, because that's going to remove a barrier to entry for somebody that just wants to get board maids that wants to play with electronics. If you are somebody that is new to this, and you are just getting started, and you send a board design out your first board design out to one of these manufacturers, and you get junk back. I mean, that is that is that's that is the first nail in the coffin to you actually wanting to do anything more with that. And because I enjoyed this so much, I want to want to remove that kind of those kinds of screw ups from people that are looking to get started or looking to start doing production runs, or the people that are established in the industry and are looking for cheap ways to to move forward. Right, because we're all trying to find ways to save money and save time.
For sure. So Dan, thank you for coming onto our podcast. Absolutely. It's been a lot of fun talking to you. Where can people find out more about you and what you do?
So I've got the two websites I've got Rheingold, heavy.com, which I unfortunately Yeah, with my last name that I chose this name for a website is in fact completely asinine. But the site is Rh e i n, G O LD heavy.com Rheingold, heavy.com or just search for you know, I squared C pull up resistors Yeah, it's it's I squared C pull up resistors. sizing and Arduino a Scylla? Or do we know oscillator how big
and then the far easier to spell manufacturing. reports.com. And then I'm on Twitter at Rheingold. Heavy and also at M net or Jesus. I should know this. I think it's at MFG reports or, my god I can't believe I don't know my own damn Twitter handle. Go to manufacturing reports.com Scroll down to the bottom look for the Twitter logo. Click on that. That'll take you to the manufacturing parts Twitter. Dan at Rheingold, heavy.com Yeah, you know, it's easy to find me. And of course this will be this will absolutely be linked in the show notes.
Yes. Yeah, thanks so much again, Dan. Yeah, thank you.
Absolutely. And you know, maybe you guys will have me back on once I go into assembly and and evaluate macro fab through a third party that will wind up having to the ordering so you don't know what's coming from exactly. I
was gonna say don't use your your real name. Yeah, of course. And with that, Dan, you want to sign us out?
Absolutely. That was the macro fab engineering podcast and I was your guest Daniel Lynch.
And we're your hosts Parker Dolman and Steven Craig. Later everyone take it easy. Thank you. Yes, you our listener for downloading our show. If you have a cool idea, project or topic, let Steven I know. Tweet us at Mac fab at Longhorn engineer or analog ng or email us at podcasts at Mac fab.com. Also check out our Slack channel. If you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest episode right when it releases and please review us wherever you listen, as it helps the show stay visible and helps new listeners find us