#243 – An interview with Macrofab - Macro Manufacturing Mechanization

Chris Gammell: This is The Amp Hour Podcast, recorded March 31st, 2015. Episode 243, with Chris and Parker from MacroFab. Macro Manufacturing Mechanization.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV blog. And I'm Chris Gammell of Contextual Electronics. And I'm Chris Church and Parker Doman of MacroFab.

Chris Gammell: That was a very professional introduction, guys. Welcome. Very nice. I would like these guys to do other things that take organization and like putting together boards for me, Dave.

Dave Jones: Really? I didn't know you were still in that business, Chris.

Chris Gammell: Oh, April Fool's? I don't know. We were talking about April Fool's jokes before the show. I don't know how we were going to work that in. We just did. Yeah, I guess that was it.

Chris Church: We need to make sure to slow his order down. Yeah.

Dave Jones: Oh, yeah. So tell us all about MacroFab. What do you guys do?

Chris Church: We are a new kind of contract manufacturer that's focused on agile electronics manufacturing for hardware startups and makers.

Dave Jones: Should people get their bingo cards out now? You've already mentioned the word agile.

Chris Church: I like the word personally. I try to fit it in every sentence.

Dave Jones: Because I see on your website here, I'm browsing through. As I scroll down, you've thrown in holistic approach and you've thrown in just-in-time production.

Chris Gammell: Oh, I didn't see this.

Dave Jones: Compliance. Well, compliance. That's important. This is all bread and butter stuff for contract manufacturers, right?

Chris Gammell: Are you guys ISO certified yet? That's usually CM's like talking about ISO too.

Chris Church: No, no. We're working on that, but I wouldn't even want to talk about that. It's pretty painful.

Parker Doman: Yeah, we have a process that's pretty similar. We're just not certified yet because that's a lot of paperwork. Yep, exactly.

Dave Jones: Now, let's get this clear up front. Do you guys actually do the contract assembly and stuff yourself or are you just a facilitator service as such?

Chris Church: We do the assembly in-house. The only thing that comes out of house right now is the PCBs themselves. Right. And we're working on solving that problem too.

Dave Jones: You're looking at making your bare boards in-house?

Chris Church: Eventually, yes.

Parker Doman: Yeah, that's the only way to get it even faster is to build the board in-house.

Dave Jones: I was going to say, does it pay you to do that or is it just a matter of speed? Just speed.

Chris Church: Speed, yeah.

Dave Jones: Speed, right.

Chris Church: Making PCBs is not a highly profitable proposition.

Dave Jones: No, I was going to say, yeah.

Chris Gammell: A lot of chemicals and crappiness too.

Dave Jones: But it's very messy. It's a very messy business. I mean, assembly is messy enough, but you're taking it to a whole new order of magnitude.

Chris Gammell: I think what you guys want to do is convince someone to open one in the space next door. That's what you really want to do. That would be great. Yeah. So anyone who's looking, Chris and Parker will help you find a space nearby, which is in Houston. So tell us, why Houston? Why are you guys in Houston?

Parker Doman: Well, this is where we're from. Yeah, I was born here in Houston and that's where me and Chris met.

Dave Jones: Sorry, I'm picturing a 44-gallon hat and a six-shooter with a quick draw.

Parker Doman: I'm sorry.

Chris Church: To be fair, they only hold five gallons.

Parker Doman: Right. And we only wear cowboy boots on weekends. Except for you.

Dave Jones: Got it. With spurs, yeah.

Parker Doman: Static spurs or anti-static spurs.

Chris Gammell: Yeah, static spurs. Yeah, static spurs.

Chris Church: Well, that's how we get conductivity to the ground.

Chris Gammell: Yeah, right. That's what they're for. Jesus. No wonder. You have like a little soldering iron on your hip. You're like, draw.

Parker Doman: And you're a six-shooter soldering iron.

Chris Church: There you go, yeah. But Houston's a great place to have a manufacturing company. I mean, if you look around, there are actually a number of EMS companies here, electronics manufacturing services. We've got about 30 in town. Really? Okay. But there's so much oil for the business here.

Dave Jones: Well, you don't have 30 in the whole country. So is it a cheap place to do business? Is it a cheap place to rent big spaces and things like that?

Parker Doman: Yeah, out of all the big cities in the States, it's one of the cheapest and inexpensive to run a business in.

Chris Church: Well, they like to tell us that. You know, they're always.

Chris Gammell: Well, I mean, the state doesn't have any business taxes either, right? Oh, no, they do. They do. Oh, they do? Yeah. Or is it income taxes or no state income taxes? That's what it is.

Chris Church: Well, yeah. What we have for the business, we have the franchise tax, which is effectively a business tax, and then property tax. So those will get you.

Dave Jones: Yeah. So would it be advantageous for your business to be in any other state? I wouldn't think so because it's all mail order business, right? Essentially, it's all sort of remote business. You could be anywhere, couldn't you?

Chris Church: Yeah. I mean, as long as, you know, we're within, you know, a couple of days delivery of any major area where our customers are in the country, I don't see any benefit of being anywhere else.

Parker Doman: Yeah, as long as the rent's not astronomically high. Right. That's only detriment to living somewhere else.

Dave Jones: Because a lot of companies, a lot of startup companies do make that mistake. Oh, we've got to go. We've got to be in California, you know. We've got to be in Silicon Valley, and they get there, and their company folds because they can't afford the rent.

Parker Doman: Yeah, the rent can cost living.

Dave Jones: Yeah. Yeah, exactly.

Chris Church: Yeah, I mean, we pay for our space here. We're in downtown Houston, and, you know, we pay probably a tenth of downtown office space in San Francisco.

Dave Jones: A tenth?

Chris Church: Yeah.

Chris Gammell: Well, San Francisco, that's kind of, that's like comparing to the highest. True, but still, that is. But that's still nice for downtown. Like, yeah, downtown, that is impressive, though. I mean, Houston is, if people don't know, Houston is huge. Like, the outer rings especially are, like, monstrous. So, that's impressive, too, that you're downtown. I mean, is there a benefit to being downtown versus out, like, in the Burbs area?

Chris Church: Well, it's kind of, it's central for all of our employees. It's also closer to my house, which is a key important thing. It's good to be king.

Parker Doman: And I actually moved out from the Burbs into central Houston to get closer to work. So, nice.

Chris Church: And we have a couple unspoken benefits, or actually, they're about to be spoken. We have a brewery right next door and a soccer stadium on the other side of us.

Chris Gammell: Nice. Oh, that's awesome. Soccer?

Dave Jones: Yeah. I didn't think you Yanks were into soccer.

Parker Doman: Oh, yeah. Actually, most of the people here at the office are pretty into soccer.

Dave Jones: Really?

Chris Gammell: I think a lot of the younger generations come up kind of playing soccer more. So, that kind of helped to boost the MLS interest. I don't know. That's my assumption. Closest team to me is Columbus.

Chris Church: Yeah, you can't really help us to be here because this whole street, they rent out our parking lot for the games. Oh, nice. All right.

Chris Gammell: Yeah, people tailgate in the front of our warehouse. Nice. Do you guys walk around selling blinky things out there? You know what?

Parker Doman: Actually, that would be a good idea.

Chris Church: Yeah, totally. A little orange solder mask on there. We could make it.

Chris Gammell: Yeah, if you guys get an extra panel, use it as your panel filler space, right? So, if you've got an extra panel, just do that. It could be cool.

Dave Jones: Oh, goodness. Now, setting up this, potentially getting into the PCB manufacturing side of things, let's get back to business. Would it be in any way advantageous to spin off a company like Nextdoor, your own company that did that, or would you keep it under the one battle? Sorry, this is just the business side of me coming in. No, no.

Chris Church: Just considering the context in which we do it, it would make more sense just to do it under the same roof, under the same business. You know, our only goal in doing PCBs in-house is to do next-day service. So, we're able to, within 24 hours, turn the whole turnkey assembly over. Right. And we would only be doing those 24-hour services in-house.

Dave Jones: Yep. Wow. And then, because there's a slippery slope here, if you set up your own PCB manufacturing, bareboard manufacturing, then you can slide down that- Yeah, it's a slippery slope. That non-profitable slope of actually contracting those services out and making bareboards. Right. To other people. And as you've said, there's not much money in that.

Chris Church: Right. Right. And then, just again, those environmental considerations here, you know, people tend to think of Texas as an anti-EPA kind of state.

Dave Jones: Yeah, yeah. That's my impression.

Chris Church: Don't mess with Texas, right, guys?

Speaker ?: Yeah.

Chris Church: But here in Houston, especially here in the downtown area, we're in a fairly highly regulated zone. All right. So, we probably wouldn't be able to do sort of the mass scale you're used to seeing. Interesting. The high chemical processes. We probably wouldn't even be able to do those here in the region where we're at right now. Yeah.

Dave Jones: And what sort of layer, what sort of process technology would you look at going for? Just like four-layer tops or would you stick to two layers or what?

Parker Doman: For 24 hours, we probably would start with two-layer only and do either lasering or CNC boards.

Chris Gammell: Oh, cool.

Chris Church: Yeah. But we would still do the solder mask and all that. We'd still have that process and have properly plated vias and everything.

Chris Gammell: Huh. Awesome. Okay. So, that's for 24-hour stuff. I mean, so can you tell us more about the current generation? Like, what is a normal turn time?

Chris Church: Yeah. So, our standard service is 15 business days and that's what we call our low-cost service. So, basically, if money is of your primary concern, that would be the service you'd choose. And we just today launched a five-day turnaround service, which we believe is still less expensive than many of the standard options you have out there. But, obviously, has a higher cost due to the need to expedite the PCBs and the parts and everything.

Dave Jones: Ah, right.

Chris Gammell: Okay.

Dave Jones: So, that's where it's big then. Now, it should be noted that you guys have what looks like – I haven't played with it much – but a fancy-pancy online – technical terms here – fancy-pancy online system that allows people to see, you know, to upload their design files, see their PCBs, like their Gerbers, how it's going to be assembled, the bill of materials, the placement, firmware, blah, blah, blah. Mm-hmm. Is this your secret sauce, so to speak?

Chris Church: That's part of it. We've actually built the technology all the way down to the line. So, not only do we have that front end there, what that does – and there's actually a step in there where you can adjust the rotation of the parts on the PCB. Right. And through that – Right.

Dave Jones: So, me, as the client, I can see that, say, somebody has goofed that part up before you actually hit the go button on the assembly machine. Exactly. So, I can double-check it myself.

Chris Church: Absolutely. And that's kind of – We actually only let you do that for the most part.

Dave Jones: Right.

Chris Church: Oh, so the client has to sign off on it before you press the go button. Absolutely. They have to approve all the parts on that.

Dave Jones: For those who aren't aware, that's not how regular contract manufacturing works. You sort of generally, like, throw it over the fence, give them all the files, and they're the ones who check it and set it up. Yeah, set up the pick-and-place machine and press go, and then you cross your fingers and hope the board comes out the other end with the parts in the right orientation and the right location.

Chris Church: Exactly.

Dave Jones: Right. So, you guys have flipped that on its head.

Chris Church: Absolutely. A lot of people don't realize, especially in the low volume, that's a huge contributor to the cost. To have a trained –

Dave Jones: Oh, set up is everything. Yeah.

Chris Church: The trained engineer to sit there for an hour and a half to read your silk screens and program the machine, rotate this part that way, if you have silk screens. Yeah.

Dave Jones: And some people complain about, oh, it cost me a couple hundred bucks for setup? Are you kidding me? It's like, well, that's cheap, you know?

Parker Doman: Yeah. I mean, for, like, when we have to set up that machine for long runs, I mean, it takes, you know, a couple hours for our operations guy to set it up. But, I mean –

Dave Jones: Yeah. And that's for a simple board, right? If you've got a real complex board, it can take all day or even two days to set up a machine sometimes.

Chris Church: If it's really complex. Okay. So, that's what we've done there is we've eliminated that upfront engineering cost by putting that back on the user. And that information you do and all that interaction you do through that user interface actually generates data for us on the back end. It tells us, for example, we know when you're changing parts around so we're able to learn where we match parts incorrectly. But also, there's adjusting the rotations eliminates that step. And, therefore, we can take the design direct to the machine. And further down from that, we still have a lot of hand assembly going on. Like, say, if you're doing a one-off of a device and you're the only person in our queue right now that has this one chip on there, we're not going to spend, you know, 30 minutes programming the machine to do it. Someone's going to do that by hand. So, we also have built an interface for our line workers where all that information flows down direct to them and it walks them through part by part on each job what to place, where, how to validate that step and everything. And that gets our cost down substantially.

Dave Jones: Did you guys have this idea from the get-go? Is this why you set up the business? Did you go, oh, look, this, you know, contract assembly sucks because you throw it over the wall. Let's build something that – or did it come about – like, did you start it and then it come about? Oh, geez, this would be a good idea. Did you slowly fall into it? How did that work?

Chris Church: It's a little bit of the former there. Previously, I had a small electronics company and we were doing very well. But we were a very niche market product and we were selling – our highest selling product was maybe 1,000 units a year. And in some products, we sell 5 or 10 units a year. And I was constantly frustrated by the fact that, you know, I put in orders with these manufacturing houses and I just don't hear anything for months.

Dave Jones: Of course you don't. You want 10 units?

Parker Doman: And even what's worse is when you hear back from them, it's some arbitrary price and – I had no idea where this bill of material just, like, astronomically is four times the price. Why is it that cost? Because we said so. That's why. Yeah, exactly.

Dave Jones: Because you guys are a nothing business to us. Go away. You're annoying. You're in control.

Chris Church: Yeah. Prior to the electronics company, I actually ran – or I had founded and was chief architect for a large-scale software as a service business doing basically cloud-based technology. And I'm sitting here going, can't this all be much better? Can't this be a much better experience? Can't we give the customer more control here? And ultimately I decided that was a lot more interesting of a problem than, you know, selling niche market electronics at the time. So I enlisted Parker. I said, let's go do this.

Dave Jones: So how did you guys set it up? Because it takes a lot of coin to set up an assembly house. I mean, you know, decent machines cost – you know, like a decent assembly loan is going to cost half a million bucks. Yes. Well, well, it certainly does here, right?

Chris Church: Yeah. Yeah. It is expensive, yes. What we did is, you know, we did – we did – we had – we've had two angel rounds at this point. Oh, okay. We raised the first in fall of 2013 and then the second in the beginning of 2014. And then what we did is we went out and scraped auctions and found equipment that would meet our needs. Yeah. And refurbished it.

Parker Doman: Yeah. I spent probably a couple months refurbishing all the machines and making sure they were running really well. Yep. They're like old cars. They just need a little TLC and just – they run fine. Nice. For a while. Yeah, for a while. And we have to do maintenance. Right, right.

Dave Jones: Was it hard to convince people to invest in a contract – a pretty much a commodity saturated market, which is contract manufacturing? Have you guys heard of China? We want to do that in the U.S.

Chris Church: You couldn't imagine how difficult it is.

Dave Jones: I'm starting to, yeah.

Chris Church: Yeah, it's surprisingly difficult. You know, I mean, fortunately though, I mean, what effectively we're building here is not a contract manufacturing company to compete with China. Right. You know, obviously, we have a much, much bigger vision. Otherwise, obviously, the investors wouldn't have had any interest. Of course. And we really see the manufacturing as a means to an end. What end is that?

Chris Gammell: Yeah, what is that?

Chris Church: Besides world domination. Yeah. Well, obviously, right? Amazon 2.0. We need a really big laser and we need money for the big laser. And the moon base. Yep. But no, what we're really focused on is what I wanted to do at my last company. I didn't want to be dealing with managing operations. I didn't want to be haggling with manufacturers all the time. I wanted to design cool products and sell them to people. So we're trying to enable startups and, you know, makers to actually bring their products completely to market. Not just the PCB, but the entire package. You know, from the software that runs on it to the PCB design to the enclosure, the packaging, everything. Getting that in the hands of the customers as easily, as quickly and as inexpensively as possible. So they can be out there designing new products and not worrying about all the challenge of actually, you know, what does it take to get this done? You know, which EMSs do I need? Which enclosure companies, et cetera? How many people do I have to call, right?

Dave Jones: Do you think this would have been as successful if it wasn't for the recent crowd funding craze where, you know, every man needs dog with an idea? Yeah. For a widget can just, you know, get a million dollars and then go, right, how do I manufacture it? And then give it to someone like you.

Chris Church: Yeah. So no one would have invested in a business like this if it weren't for crowd funding.

Dave Jones: I was thinking that.

Parker Doman: Yeah. I would say actually crowd funding pretty much, it really helps the small guys get their product out there. And really enables a company like us to try to exist.

Chris Church: And on top of that, it's brought hardware into the mainstream up for the investors. Oh, exactly. The public. Yeah. Now a little guy can bring a product to market. If they try really hard. With the help of you guys.

Dave Jones: Yeah, exactly. Can they, how many of your customers, you know, are the panicked, oh my God, I just got half a million bucks on Kickstarter. I don't know what to do.

Chris Church: Help. Usually we're, we catch them before they get there.

Dave Jones: Nah, right.

Chris Church: Yeah.

Dave Jones: So do you proactively go out and search crowd funding campaigns and then email and then spam them?

Chris Church: No. No.

Dave Jones: As I've gotten several times?

Chris Church: Yeah. Dear sir. We kind of make a point not to do that because I just, you know, it's not a very cost effective strategy. We'll put Dave on the mailing list. Yeah.

Chris Gammell: Dave, do you make Bluetooth speakers by chance? Because I always, maybe you want to make a Bluetooth speaker. Have you gotten those emails?

Parker Doman: Are the worst lists of features and they say how much? Yeah. No. And they're like, uh.

Dave Jones: Oh, have you ever turned down anyone because you think that they're just scammers or their idea is just so shit it's not going to work and you don't want to be involved in it?

Chris Church: We don't turn them down. Um, generally what we do is we, uh, we, we, we like to use what we call, uh, uh, price, uh, sort of priceability in this. So anything that we don't want to do, we just price it at a price that someone else doesn't want to pay. Yeah, right. That seems to be effective. It's better. It's better than saying no.

Parker Doman: Right. Right. I mean, I, I don't, I don't want to turn away people and I'd like to help people in their designs and stuff. So.

Dave Jones: But some people are just idiots and need to be safe from themselves. Right. That's. I wouldn't put it that way, but. You're correct.

Chris Gammell: There's a high cost to involving yourself. How about you say it like that? Yeah. That's. That's. Yeah.

Chris Church: So we, we just charge what it's worth in that case. And usually we find it's not worth that much to them.

Chris Gammell: Yeah. That makes sense. So, I mean, uh, Dave's asking about the kind of the prototype quantities and it seems like this is very, very strong. Like probably one of the strongest, you know, I think about you guys and then small batch and circuit hub. That's kind of like the, the group of like assembly with like some software front end versus kind of the traditional, the, uh, like advanced circuits and those guys, like all the Sierra, all them, like everybody, you know, like those are kind of the, the, uh, the options I think of, but I, I actually looked at the price on the, your, your example. And I, I do recommend people go to macrofab.net and then click on the, uh, what's the button? Is it the, it's the demo button at the, at the top. And basically it loads up, uh, a propeller development stick board, stuff like that. That was actually what sold me on it so far. I mean, I, I haven't had a chance to try it yet, but just the single board costs assembled for 52 and a half dollars. Yeah. That's crazy. Insane. Uh, now I then tried to drag it out to kind of more quantities. What are you guys, do you guys have a, uh, a break even point where you say, well, at a certain point, it doesn't really make as much sense for us or, you know, then we have to incur different costs. Like how, how do you guys see your split between single quantity, a hundred thousand million?

Chris Church: Right. So with that, uh, generally at a certain point, everybody hits the same level of efficiency at a certain volume. Um, and so what we find is that once you get into several thousand units in a single order, our pricing now starts to creep up enough relative to other, you know, small assembly companies that a lot of times it doesn't make sense to the customer. Um, right. Generally, we want customers that are below a thousand units a month. Um, that, that's really our sweet spot. Uh, our software is much better at scheduling, uh, when we can get a job done in a few days, uh, versus, you know, dragging it out for, for weeks. Or in our case, you know, adding another second assembly line to get a job we may not get a repeat of is not necessarily very good for the business. Um, but we, we find right now, you know, our primary goal in that was to get those single unit turnkeys as, as inexpensive as possible. Um, now that we have that, right, and we were able to actually achieve that and we're doing that profitably on our side, now we're going to start looking at, you know, how, how we can continue to extrapolate those, those savings up, up the, the quantity change.

Dave Jones: Yeah. Now, one of the key parts of your ability to offer a service, turnkey service like this is the fact that you've got a massive, well, a large array of, um, house parts, right? And if people are using those house parts, which are already loaded on your machines, I would presume, then the cost is going to come way down.

Chris Church: Right. Uh. Is that right? Yeah. So what we, we do two things with that, uh, from the cost perspective. One is that for our house parts, we offer those at a lower cost than you can typically get those in the volumes you're looking at. And then secondly, we give away the labor for free on those. Right. Because as you've noticed, noted, we already have the machines trained. I do want to jump back to one of those, uh, keywords we used earlier. We were talking about the bingo. I'm going to use it now. Uh. Right. It, it does also assist with just-in-time manufacturing and the predictability of manufacturing. Yeah. And then we always guarantee those parts will be available. And if we're going to end the life in any of those, we give you a personal six month notice on that. Oh. Got it. Nice.

Dave Jones: Now, at the moment, you've only got passives, haven't you? There's some passives. Well, there's some diodes and things like that, but you don't have any chips as such. Yeah.

Parker Doman: It's pretty much a jelly bean parts. Um. Yep. Um. Resistors, capacitors, that kind of stuff. Connectors. Um. We're bringing in some more connectors, I think. Did you, did you do the block with you, Adron? No, no.

Chris Church: We're going to do the, uh. This is not an April Fool's joke. We're going to have a, uh. Someone hit the mute key. Mute, mute, mute. Yeah. We're going to have a new set of house parts, uh, a new addition to the house parts list, uh, coming out around Monday. Cool. Nice. Yes. And.

Dave Jones: And people can download these in some usable.

Parker Doman: Yeah. So, right now we only have, um, Eagle footprints of this stuff. Um. We're working on KCAD and DipTrace support. But the, the data sheets are all there for the parts, so.

Chris Gammell: Yeah. Yeah. Right. So, you just match the part number of what you guys have. Mm-hmm. The manufacturer part number and it should just pull it right in. Mm-hmm. Yeah. Nice.

Chris Church: And a lot of cases will, like, if you're putting in resistors and capacitors, will automatically match based on the value and the footprint. Oh, nice. We'll automatically match house parts and suggest those to you.

Dave Jones: Excellent. Now, I noticed that you go down to 0402. Mm-hmm. Is, you had too many issues with 0402 placement? Not, not yet.

Parker Doman: Oh, okay. Our machine's actually able to go down to 0201. We haven't tried that. Yeah, that'd be. It says it can do it.

Dave Jones: There's a big difference between what it says it can do and what yield you get in practice.

Parker Doman: So, we have a universal GSM2 and that's a mid-90s machine when it was designed. I think ours was produced in the late- Late 2000s. Mid-2000s. Yeah, mid-2000s. And it can do 0402 pretty well. Right. That's at the threshold of its vision system that we have currently set up. We have to upgrade its cameras to do 0201.

Dave Jones: Got it. And have you only got the single line or have you got multiple lines?

Parker Doman: Yeah, there's only one line.

Dave Jones: Okay. Have you got any, at what point do you expand to a second line? When it's filled. And what sort of throughput business, and what kind of increase?

Parker Doman: So, it can do, our current line can do about 6,000 parts an hour. So, once we're at that, we'll probably start looking at getting either another machine to add on to the front end, like a chip shooter or just add on to another line. Yeah. And we can always add chips.

Dave Jones: What is the bottleneck? Is the bottleneck? Because I assume that the bottleneck is like just the manual handling and dicking around and things like that. It's not actually the speed of the machine. I would be, especially if you're doing lots of, like if you've got 10 different jobs to do today, I don't think the speed of the machine would come into it, right?

Chris Church: It's going to depend on which processes those jobs have. Because the pick and place, say, if we're doing a lot of chip parts, I mean, like leaded chip parts, that can slow down quite a bit while it's validating those. But if it's a whole bunch of resistors and passives, it moves pretty fast. But it's going to depend on the mix of the job. Sometimes just going through the reflow of it is the slowest part of the process.

Dave Jones: Ah, yeah.

Chris Church: It takes about six, seven minutes minimum.

Dave Jones: It's painful, isn't it? Yeah.

Parker Doman: I'll say the worst is actually, because we have a selective solder to do our through hole parts. And actually trimming the leads for all the parts before it goes in, that's what takes the longest. Yeah. Because we build a pinball board for a company. And cutting all of the mosset leads and transistor leads, that takes like 10 minutes. And the surface mount goes through in about four and a half minutes.

Speaker ?: Yeah.

Parker Doman: Can you guys get like a bandsaw for that or what? Why don't we just manually clip them?

Chris Gammell: Yeah.

Dave Jones: We'll probably be adding a lead former later. How many people have you got working there maintaining the line, if that's not corporate?

Chris Church: So we have four full-time employees right now. And our line workers tend to be based on capacity.

Dave Jones: Oh, okay. Right. So they're just sort of casual. They come in and out.

Chris Church: Right. When we need them, we call them in. But usually we can handle a pretty decent amount of volume with actually just two people working the line at any given time.

Chris Gammell: Wow. Nice. Yep. So for the setup stuff, because I'm still a little confused about this. Okay. So first question, I think, is how many inline feeders do you have?

Parker Doman: So the machine can do 64 feeders. Plus the PTF. Yeah. Plus the PTF. So it's 32 per side. And the PTF is the platform tray feeder, which has 28 slots in it. So you can put 56 tray parts in there. And that basically has an elevator that indexes the parts. And it pulls the tray out, picks the part out, puts it on the conveyor, and runs the conveyor into the pick and place. And the pick and place picks it off the conveyor and then puts it on the board. Oh, cool. That's impressive. Yeah. So we can run a lot of IC parts if they're on tray.

Chris Church: And we can actually mix trays too. So on one tray, we can have 30 different parts on that tray.

Chris Gammell: Yeah. Huh.

Dave Jones: But so would you prefer people order chips on trays or would you prefer they, like, you know, a standard SO8 or SO14, would you prefer they're on reels or on trays?

Chris Church: We would prefer they let us order them for them.

Dave Jones: Right.

Speaker ?: Yeah.

Parker Doman: So depending on the size of the part, if it's small, if it fits under 12 millimeter tape, we prefer it on tape. If it's over that, then the PTF seems to win out in terms of speed and optimizing on space. Because the wider feeders will take up more space on the machine.

Chris Gammell: Ah, okay. So the other question I have about that then is, so do you, so right now that you have a single board type thing, does that mean that you run mixed panels like Oshpark does where there's multiple boards per panel? Yeah. Exactly. Ah, okay. Isn't it possible, though, that, like, I'm imagining, you know, like an 18 by 24, you know, not like a huge panel. I don't know how big the panels you guys are using is, but, like, if you had, you know, maybe 30 different designs and all the designs are radically different, which probably isn't the case, but if they were all radically different, isn't it possible that you have to run multiple times through with the feeders or how does that all work?

Chris Church: Yeah, so if we're talking about 30 different one-off designs on there, the only parts that are going to be on feeders are going to be the house parts or common parts that we're likely to have in stock. And in those one-off designs, we're actually likely to place many of the unique parts on those by hand.

Chris Gammell: Ah, okay, okay.

Chris Church: Because it's faster than trying to load up deals. And the wastage is, once you account for the wastage and trying to reel up those parts and then the additional labor, it actually gets more expensive to try to machine place those than do the by hand. Okay.

Dave Jones: How difficult does warehousing and sorting of parts actually become, you know? Because I've been to these pick-and-place houses and they have just walls and walls of, and they have their own system for, you know, stock and all that sort of jazz.

Parker Doman: That is the hardest thing. Oh, really? That is the hardest thing about my job. Yep. I probably spend half my time cursing at our inventory system. And he's looking at me because I wrote that inventory system.

Chris Gammell: Do you guys have, like, vision for it or how does it work? Is it just kind of indexing? It's indexing and barcode-based at the moment. Okay. Yeah.

Chris Church: Yeah. And it's fully, and again, it's integrated. So when someone places an order in, you know, we quote the parts for them and everything, and we actually see all the jobs coming through and we see what needs to be ordered, what's in inventory, you know, what distributor to order it from. So it's a fairly automated process on our end, but it does require a lot of data entry to keep it up to date.

Dave Jones: How do you physically store all the parts? Do you have, like, pigeonholes on the wall or do you have one of those fancy-pantsy automated machines that sort of you punch in the number and it rotates its carousel and it spits out the correct parts?

Chris Church: You know that thing earlier I said where we buy old machines and refurbish them? Oh, right. Yeah. We don't have anything fancy-pants. So boxes? Yeah. Right now we use protective bins and we're actually building out about 150, 200 square foot humidity controlled room right now.

Dave Jones: Oh, wow. Nice. Okay. What parts have to go into the humidity controlled room?

Parker Doman: Pretty much, we're going to put everything in there because we'll have the space. Right. Okay. Because if you have a house part reel that's on paper that just lies around for a while, in 60% humidity that paper is going to stick to that tape eventually. Oh, Houston, right. Yeah. And so that tape is just never going to peel off right and you're going to have jams in your feeders. So keeping just legit parts dry is pretty important for longevity of the material that holds the parts together. But LEDs do in particular hate the humidity.

Dave Jones: Yes. Yes, they do. Because when they heat up, if people don't know this, when the parts heat up, they can get moisture inside them. And then when they go through the reflow oven, that moisture inside can expand and then crack, you know, the lead can just, you know. Oops.

Parker Doman: Yeah, that's the problem with when people ship us parts and we just get like cut tape of like a FTDI chip that's highly susceptible to moisture. And you're like, I have no idea how long these have been on someone's shelf.

Chris Gammell: 1977 date code. All right. Yeah.

Chris Church: And, you know, traditionally we'd have to, anything that's out of spec on the humidity, we have to pre-bake it before we can start placing it on the line. And that's just a nightmare. We want to minimize that as much as possible.

Dave Jones: Yep. That's annoying. Have you ever been bitten by fake parts? Have you ever been duped?

Parker Doman: Not yet. We try to buy everything through authorized distributors. And we pretty much hammer that home when people say, well, I can get this part cheaper on AliExpress. And I'm like, oh, good luck.

Dave Jones: Yeah, good luck. Yeah, exactly. You can hand side of that one. The full week lead time from China, you know. Yep. So I can imagine with your specialty being fast turnaround, you'd be a big user of Mouser, DigiKey, and all the other, you know, turnkey stockers. Yeah, we are.

Chris Church: We are.

Dave Jones: Is that where you get the majority of your parts from? Yeah. It seems to make sense, you know, and then you build that into your cost.

Parker Doman: Yeah, the majority actually goes through Mouser since they're here in Texas. And we actually get ground shipping next day from them. Nice. Sweet.

Dave Jones: Because people think that, you know, Mouser and DigiKey and the like are like super expensive, but they're not. You go and buy a reel of parts, it's not going to be much cheaper from an official distributor than it is from Mouser DigiKey. Well, unless you get like special pricing.

Chris Gammell: If you get pricing direct from like TI or Maxim or something. Yeah.

Dave Jones: Right. But generally speaking, it's, you know, for your average punter who can get hold of these parts from your regular distributors, they're not that much more expensive. Is that what you guys have found?

Chris Church: Yeah. In fact, one of the things that really eats you up when you're dealing with all these different distributors trying to find the cheapest possible prices for the parts are the shipping. A lot of people don't take that into account. They're like, oh, I can get two of those parts for 30 cents each from this other distributor. So you should sell them to me at 30 cents each. And I'm like, but yeah, but it's eight bucks in shipping.

Dave Jones: That's dirt cheap. Some people, some people insist on using Kuri and charging you 50 bucks. Right. You know? Well, over here in Australia you do because you've got to get them shipped, you know, internationally. I guess it's a bit better over there, but yeah.

Chris Gammell: Well, you guys have to put that stuff on a shelf though, too. That's the other thing that people probably don't realize. It's just like the handling costs of anything. Like anytime you click add a part to your bomb on Macrofab, I'm sure that there's some overhead there just from processing, right? Yeah.

Parker Doman: Inventory, handling, making sure it goes in the right bin, et cetera.

Chris Church: Well, yeah. And at the end of the day, like any other good EMS, we have to provide lock control. So that means every part that comes in here, if we order one resistor for you, right, we have to record what lock that resistor came out of. And a human has to sit there and record that information. Oh, my God. Yeah.

Chris Gammell: That's not integrated. There's no like, there's no digital integration for that kind of stuff?

Chris Church: We're working on that. It's sort of a, you have to do it from scratch with every distributor and it takes a while.

Dave Jones: Wow.

Chris Church: Trying to get them moving to integrate with you, who they think is a fairly small player. There's a lot of times like, you know, kind of watching molasses come out of the tree. Yeah. They'll get to you next week. Then they'll tell you a time when they're going to get to you again to tell you when they might actually get to you.

Dave Jones: Yeah. That's it.

Chris Gammell: You whippersnappers and your bits and your programs and your Python. So actually, I've worked with manufacturers that don't do that, don't do lock control at all. So that, I mean, that is, that's a nice value add for people that are using your service. That's good.

Chris Church: Yeah. I mean, because, you know, at the end of the day, we're going to charge you to manage your bomb for you. So we should at least provide a good service, right?

Chris Gammell: Yeah. Right. And so do you guys find that you're, you're doing a lot of the, so, I mean, you have the in-house parts. Do you guys ever make the decision, like, so say I go order an op amp, will you just order like a cut tape of like, like a 10 cut tape of that just for my project? Or will you guys sometimes then go and say, we're going to order that whole reel and then keep it in-house? Like, how do you, do you make that decision? And if you do, how do you make it?

Parker Doman: Uh, it really depends on the part, um, and how many you're ordering, um, and how the price breaks of that part work out.

Chris Gammell: Oh, okay. So you do like a break-even analysis and then say, if it's above this, then we might as well just have it as a reel.

Parker Doman: Exactly. Um, and there's also like, if you're ordering just one board and if it's not a rush job, we'll usually not even order overage on a chip part, like an IC. Yeah. We'll just get one. Um, yeah. Which is kind of a pain in the butt to inventory, but, um, it, it works out. If it's a rush job, we pretty much always get overage.

Dave Jones: Would you, would you even bother? Because you're just getting the one part and you know that person's only ordered one. Would you even bother at that point? Wouldn't you just get the part and whack it straight on the board?

Chris Church: Well, it's, it's, in some cases, you know, we're able, we see not just the one order. We see all the orders that are coming through the system. And we also see the projects that aren't orders yet.

Dave Jones: Ah, right. So you can see when, when people are trialing on the website and uploading there.

Chris Church: And we can, and we can estimate demand from that. So it's not unusual for us to look at this and go, you know, Hey, this part over here, uh, you know, 15% of the projects that have been uploaded are using a part that this would serve, you know, function for. So why don't we carry it in house?

Parker Doman: Yeah. It's like the, actually the part we're adding soon is the, uh, like a bat one 48 diode. Uh, and that's 42, right? Oh, it's bat 42. Bat 42. Yeah. The jelly bean diode insert number. Um, so yeah, we're adding that to it because we see that a lot in our, in our projects that are being uploaded.

Chris Church: And we've seen it as much as like 20% of all projects, not necessarily the bat 42, but that or one of those class have been in there. So. Yeah.

Chris Gammell: Hmm.

Dave Jones: That, that'd be rather frustrating because you see people, you know, using all these subtly different parts and you know, they could consolidate, you know, they know that, oh, why didn't they use a BC 547? They chose a, you know, some other one, which is, you know, it's just doing some jelly bean switching operation. They could have used a 547, but they go and use some other weird ass number. Right.

Chris Gammell: We just got to charge 50 bucks for that one, right? I mean, you do it economically.

Parker Doman: Yeah. It has a placement charge at that point.

Chris Church: And that's one thing we actually do, especially with the resistors is we try to match within ranges. You know, you'll, you'll, you'll, people will put exact values for the resistors because they're there with the calculator. And when you're doing the schematic design, you're like, yeah, that needs to be a 6.317 Ohm. That's what I'm using. Right. And you can only get it on 20 week lead time. Okay. So we'll, we'll look at that and automatically if someone doesn't put in a specific part number, I must have this part. We'll start suggesting things within a certain range of the values they provide so that they can kind of see, hey, maybe this will work for all of those components you have there.

Dave Jones: All right. Is that sort of like a semi-automated online, not automated, but like, like you don't have to actually send, do you have to like send an email and go, you know, hello, Mr. Jones. You know, we found that you blah, use this part, blah, blah, blah. Or do you, is that all part of the online tool? It does it automatically. Exactly. Which you've got.

Chris Church: When you upload your files, as it's processing the building materials, it does it automatically.

Dave Jones: Ah, excellent. Ooh, I'm going to have to try this. This looks really jazzy.

Parker Doman: We've actually been finding a lot of users will, once they realize there is a house part list, they'll actually change the design and use these parts. It also helps that we do have footprints for Eagle. Definitely need to get the other EVA tools. Yeah. Yeah.

Chris Gammell: Yeah.

Dave Jones: Is that the holy grail of everything? Like having so many in-house parts that, you know, people would rarely have to do something different, you know?

Parker Doman: That's going to be the end goal for the house part list is we'll have so many well-used parts that we don't have to really worry about.

Chris Church: You know, weird parts.

Parker Doman: Yeah, yeah.

Chris Church: And then, you know, there are just the parts that are truly unique and custom and aren't used in every single project out there.

Dave Jones: Which seems to be every one of my designs because I'm always doing something niche, you know? I'm using some, you know, weird-ass part that nobody else has ever used. Yeah, but you don't use all of that. You use cast-less as well, right?

Speaker ?: No, no, no.

Dave Jones: It's, you know, yeah, exactly. Yeah. You know, 90% of it might be, you know, Joe Blog stuff. But, yeah, then there's like, you know, 10% of these weird-ass parts that, oh, I can only get from one. Only one person in the world's got stock of, you know?

Chris Church: Yeah, actually, we've had a few designs like that recently. And one guy, it's a one-time programmable chip. So they had to be sent to us pre-programmed and we couldn't just order them for them. Yeah.

Chris Gammell: Right. It's got that crazy part number that nobody understands because it's got the one, the OTP number that's assigned or whatever, right?

Dave Jones: Do you order your microcontrollers pre-programmed because Mouser and DigiKey offer and the manufacturers and distributors offer that service to pre-program the micros? Or do you do it yourself?

Chris Church: We're, right now, we're doing it ourselves, but we are looking to switch to that where appropriate. Right. Because one of the things we're going to be doing, the feature isn't fully enabled in the UI yet with the firmware. You're actually going to be able to specify multiple different pieces of firmware and software to upload if it's part of a multi-step process. And then we're going to leverage that same capability from the front end to capture the testing process.

Dave Jones: So this will be especially useful if people haven't put an in-circuit serial programming header on their board, right? If they're using like a little, you know, a SO8 or even one of those little SOT23 micros and they haven't, you know, they're using all the pins, they haven't built in a programming port. Well, how do you program the part? Right, right. You know, well, you may as well just order it pre-programmed from the distributor. Send them your firmware file and be done with it.

Chris Church: The only challenge we've run into with that is getting them to negotiate on volume pricing.

Dave Jones: Oh, okay.

Parker Doman: Yeah, so for low volume stuff, it makes way more sense for us to do the programming in-house. Right. It's when, you know, they need 2,000 of these things programmed.

Dave Jones: Even for those fiddly parts, what if somebody sends you a little 5-pin SOT23 micro? That's real fiddly to put in like a zero insertion force test thing.

Chris Church: It may, in that case, make sense to order it from the manufacturer. I mean, from the distributor pre-programmed.

Chris Gammell: Yeah, well, they're doing the same thing, right? I mean, you got the fiddly fixture, they do. Right, right. Right, I know. Somebody's got it.

Chris Church: It depends on how bored Parker is and whether or not he wants to build a fixture for that today.

Dave Jones: That's great. I'm surprised you do this with so few people, actually. But I'm guessing you guys are young, single, and you don't have lives, right?

Chris Church: Well, a couple of us are.

Dave Jones: Right.

Chris Church: No, I mean, at the end of the day, it's really about effective time management. We have more people, full-time people in engineering than anything else here. We look at myself, we have another full-time developer in addition to me. Parker's primarily in engineering. And then we have one full-time ops guy. And then we backfill that with labor as we need. Because, you know, effectively the lines, until you get to a certain capacity and you're just, you're having to just manage a lot of the people and the tasks that have to be done, they're pretty efficient at this point. And with the stuff we've built, you know, there's not a lot of question of what we have to do on any given day or how to do it. And it's just a matter of do we have enough hours on it today.

Dave Jones: Do you have any, how many projects would you get in that are just, people are just so clueless at either board layout or design for manufacturing?

Chris Gammell: I was going to actually ask, how crappy can my bomb be? So if we could throw that one in there as well. Yeah. If I could be the example crappy uploader.

Dave Jones: More to the point, what percentage of projects would come in that are just so easy for you to process and do and that are all right and the people obviously knew what they were doing compared to the ones who don't.

Chris Church: So we're, at this stage, we're young enough, we're nimble enough that we like to say we're dedicated to every customer's success. So we'll do unnatural things. Oh, tick. Hang on.

Dave Jones: I'll just cross that off my jingo sheet. Correct.

Chris Church: But no, we'll literally do unnatural things like sit there for two hours on the phone with the customer, helping them correct their bomb, doing it for them in the UI. If they run into issues uploading, we go and just do it ourselves for them. Yeah. And get it all prepared. All right. But actually, a huge chunk, I want to say at least 70% of the projects we see are at least the customer can themselves get it to what we would call a functioning product within an hour on the site.

Chris Gammell: Wow. Nice. That's very impressive. Awesome. Yeah.

Parker Doman: Yeah. Currently, Chris was talking about the bomb. We currently have a specific format that has to be – so we have all the information. We need to display all the information and have all the lines set up for that job. Yeah. We are going to have like an importer eventually, but –

Chris Church: Yeah. Well, I mean, to be clear, for certain CAD tools, we can extract that automatically. Yeah. Eagle, dip trace, key CAD. We can do that automatically from the source files. But for other EDA tools, we don't just take any old garbage throw at us, if you will, right? It has to be a specific format. And if they can make it past that hurdle, they're usually good at that point.

Dave Jones: Right. Well, that was going to be my next question. What if I don't use Eagle? Can I just upload my Gerbers in a text bill of materials? Yeah.

Chris Church: So you can upload your Gerber files. And we actually – we don't call it – we call it the X, Y, or S file, which is the merge bill of materials and placement data. So – Oh, okay. We need both of that to actually manufacture it because, obviously, we can't automate the assembly. Of course. We don't know where the parts go.

Chris Gammell: So there's like a – there's a format that people can just go and match against like in Excel or CSV or whatever. Just this column here, this column there. Not an idea. Absolutely. Huh. That is very cool. Yeah. And especially – I have to try it, Chris. Well, yeah. So I have tried – I have tried it with the BenchBuddy. And actually, I did want to ask about that. Can you guys – so does it not show the fourth layer? It's a four-layer board. Wait. Does it not do four-layer right now?

Chris Church: We don't do four-layers through the UI yet. Okay. That's coming very soon, though. Okay. We can, though. We are doing four-layers for a number of customers. All we ask right now is you just tell us that you need to do it and we'll adjust your quote before you buy it.

Chris Gammell: Oh, just because PCBs are pricier kind of thing?

Chris Church: Yeah. Yeah. Yeah. It's substantially more expensive than the very small volumes.

Parker Doman: Yeah. The four-layer is going to come pretty soon. I mean, I think it's – we're within weeks of that. Yeah. Probably now.

Chris Gammell: Cool.

Parker Doman: Yeah. I mean –

Dave Jones: Who's making your bare boards at the moment? Are they local?

Parker Doman: So it depends on the volume. For one-offs and fast stuff, it's actually Osh Park does our boards.

Dave Jones: Oh, right.

Parker Doman: So for the higher volume and the big panels, we go over to – we have a shop over in Korea that we have a broker with. And they do a pretty good job on the boards. The quality is pretty good.

Chris Church: The price is good. They're consistent on their delivery. Yeah. We've had a lot of bad experiences. And when we do – we are happy to shop around as we're doing that. A lot of times, you know, depending on the number of one-offs we get in a day, they'll all go to Osh Park because they're just not a panel full worth of one-offs. And some days, you know, we've got – we cross past our panel threshold and they go to Korea.

Chris Gammell: And so when they do go to Osh Park, do they – so that's just all hand placement because it's not actually going through the machine? Or do you have, like, brackets that hold it when it goes to the ticket place?

Parker Doman: So we have a jig that will hold it when it goes to the machine that's indexed so that we can – so the machine knows where the board's at. But there's also a lot of hand placements too on some of that.

Chris Church: Yeah. And again, when we're looking at a one-off, it may sometimes just make sense just to do it all by hand.

Chris Gammell: Yeah. Yeah. And I know this is probably a sensitive question, but are the prices real? Like, it's not like – it's not like a – because, I mean, okay, to be honest, like, I could very easily see that low-quantity stuff, like, if you were interested in that, you could make that a loss later, right? Because then you're trying to get people in and trying to do that thing. I mean, so I don't know if you guys want to talk about that, but is it real? Yes or no is fine.

Chris Church: So as long as we don't make a mistake on any job, those prices you see there are profitable for us. Perfect. We don't – And that's what I'd expect. And I'll be honest with you, right? We're not going to grow and be a huge company on one of the prototypes. Right.

Dave Jones: No, no, no. Exactly. But I see that – like, I'm typing the quantity in here. Like, for that example board, like, it's $52 to get one board manufactured, assembled. Does that include shipping? No, the shipping is determined.

Chris Church: No, right. You get to that point once you put in your shipping address.

Dave Jones: Okay. It's $52 for one board. Then if I type in 10 boards, it drops down to $45. And then I go up to 100 boards, and it drops down to $34.80. So it's not – you know, it's not like a massive sort of, you know, drastic drop. And if I go up to $1,000, it's $22.70. People might think, oh, if it's $50 for one board, if you're $58 or whatever it was, then, well, it must be $5 for 1,000 boards. But it's not, right? Right now.

Chris Church: Yeah.

Dave Jones: It's not quite to that level. And what is the bottleneck in the pricing there? Is it the labor? Is it the parts? Is it the –

Chris Church: It's mostly – Bear board, what? It's two things. It's the PCB and the parts themselves.

Dave Jones: So the bare PCB, so the raw material costs.

Chris Church: Yeah.

Parker Doman: You can't get lower than raw material costs, so –

Dave Jones: Of course.

Parker Doman: Yeah.

Chris Church: And honestly, we, you know, we are a for-profit business, so we can't give away, you know, the labor we have to do. You know, we can't just say we're going to charge you exactly what we pay for the PCBs and, you know, lose money on the shipping and everything else. So –

Dave Jones: Well, I noticed that for the one-off board that costs you 50, costs the customer 50 to 50, I mean, $12.22 of that is labor. I assume you're only – are you only factoring, like, you know, generic running of the assembly machine kind of labor?

Chris Church: So, in a one-off, actually, what we do is we look at the actual parts you have on the board and the kinds of parts they are, how they are mounted on the board, and we charge for a one-off the worst-case scenario, effectively, if we have to hand-place each part. That's what we charge on one-off. That makes sense.

Dave Jones: But it's $12. That's the thing that I guess Chris and I are shocked at. It's $12 for labor. That doesn't even pay for one hour of a low – you know, sort of a, you know, a low-end worker. Right.

Chris Church: It shouldn't take them one hour to put that board together. If so, they're way too slow.

Dave Jones: Okay. But then – but the one-off board, there's all the stuff in the background. You guys have to do work there as well.

Chris Gammell: Yeah, but there's a lot of placement. Dave, you're looking at a board that uses in-house parts, though, too, right? This has parts on it that are – Oh, right. Yeah. Okay. Yeah, so all – So you'd have to upload a different one.

Parker Doman: All those resistors and capacitors are already on that machine. So it just rides on through that picking place and gets populated. And then I think it's like the SD card and a couple of the – Some of the chips. Yeah, and some of the chips just get hand-placed. So it's like, you know, you're talking like six parts just have to be placed manually. Yeah. And that takes about – Which is fairly quick. Yeah, it's about 10 seconds per part.

Chris Church: Minutes.

Dave Jones: Right.

Chris Church: And you'll see if you look at like the – If you go to the bomb viewer there and you kind of scroll down to one of those chip parts, you'll see we're actually charging you for – You know, to hand-place one of those microchips about, I think, like the pick. We're charging like three bucks. Oh, yes, I see it. Just to place that. It's the prop. Yep. The prop, sorry.

Dave Jones: Oh, I see it. Yeah, it's individually. There's one part here. It's to mount the six megahertz crystal is going to cost me 52 cents in labor. Right.

Chris Church: Wow.

Dave Jones: That's incredible.

Chris Church: And then you can optimize for that stuff. And then you can optimize for that stuff. And then you can optimize for that stuff. That's the volume that labor charge scales. Yeah.

Dave Jones: Right. Yes. Got it. Got it. Very cool. So that's the one of – so the propeller chip, the 44-pin quad flat pack, that's $3.30 to actually place that. Wow. The visibility in –

Chris Gammell: Yeah. This is what I always wanted because – yeah, because when you're optimizing for a manufacturing process too, you want to – like engineers want to know this. We want to know what knobs can I turn. And if I go replace the part with a different footprint, maybe it takes up more space and it costs a little bit more on the PCB. But now I trade off that little bit more PCB space and I save a bunch on the placement stuff. So yeah, that's really awesome.

Chris Church: Yeah, it's amazing. And we're even working on making that easier because right now you have to find the part in the bill of materials, adjust your volume and all of that. We're actually adding a feature soon that will allow you at the searching phase where you're going to be searching for a part to put in your bill of materials there. It's going to give you a full comparison on all of those axes of the price difference at different volumes. So you can see, for example, a house part may be cheaper at a very low volume, but a particular market part, once you get to 1,000 units, that one actually saves you $0.03 a board. And be able to see and visualize that out so you can make more informed decisions about what you're purchasing.

Chris Gammell: Yeah, that's awesome. That's really awesome. I love it. I want to ask –

Dave Jones: It's just so jazzy. The visibility is just like – Dave, you should try it, yeah. Yeah, I'll get my – I'll get Dave 2, who's my other – Mini Dave. He's my mini Dave. I might get him to put my microcurrent through this as a test and see what spits out the other end, maybe.

Chris Church: And one thing, when you're in there, we get a lot of questions about things a lot of time. We really wanted to have this to be a good user experience, so there's not a lot of labels for everything. But almost everything in that screen is clickable, or you can mouse your mouse over it, and it gives you more information. Yeah, that's nice. Just start clicking on stuff, and you'll see the whole thing will start to change and rearrange based on what you click on.

Dave Jones: Yeah, I can see it. Click to edit everything in the bill of materials. Wow.

Chris Gammell: That is very nice.

Dave Jones: Very nice. I'm impressed.

Chris Gammell: I wanted to ask about the – how do you guys do solder paste as well? Do you have one of the solder paste printers, or do you guys actually stencil?

Parker Doman: Yeah, so right now we do stencils. Low volume stuff, it's captain. High volume, we go stainless with a frameless stencil. I really, really, really want a jitter. Yeah. He's working on me every day about that.

Chris Gammell: Yeah, yeah. I'm on me every day.

Parker Doman: Yeah, they're not cheap. I play with one that's like 20 grand. Oh, yeah, that's it?

Chris Gammell: That's actually a pretty good price. Yeah, it is.

Chris Church: It's not going to work for 10 grand.

Chris Gammell: Oh, it totally will. Eventually, yeah.

Dave Jones: But they're not that quick. They're pretty slow. Yeah. They're generally more for rework than they are for – well, the ones I've used are generally more for rework purposes. Oh, dude, I've seen some fast ones.

Chris Gammell: Yeah, so –

Parker Doman: I saw one in Michigan that was insane. Yeah, we – I've got to see some before, and even ones that are moderately speed, like – if they can do like 2,000 dots an hour, I mean, that would actually work pretty well for us at the current moment. All right. And we could bump the price a little bit down since we don't have to get stencils at that point.

Chris Church: And I keep telling him, you know, we could give away the stencils for free for like two years for the price of one of those.

Chris Gammell: Long-term investment, Chris. Just Parker, keep working on it. Yeah, I'm working on it. Just keep – you can do it, man. Just whatever the cost-benefit analysis. Just keep doing those and slip it under his door or mail them to his house.

Chris Church: At some point, they'll tell me that that's the primary bottleneck in the line, and then that's when we'll get one.

Parker Doman: Yeah. Yeah, that's actually – yeah, yeah, yeah. And that's what actually takes the long – on the stainless stencils is swapping the stencils out. Because we have a semi-automatic machine, which means we have to load the board manually, and then you press print, and it goes and wipes the board, and then opens it back up. Setting that up correctly is the – that takes longer to set that up correctly sometimes than the actual pick-and-place machine. So having a stencil printer would actually make that much faster. Problem is we have to give it more data from the users to make that process work.

Chris Gammell: Right. Right. Oh, you mean if they were not using the – yeah, because I mean right now if they pull in the design files, that's not a big deal because you can just extract that stuff, right? Yeah. But –

Parker Doman: Actually, we get most of the data we can get from Eagle, KiCad, and DipTrace. We can get the paste – so if you look at like a paste file, it'll show the apertures of where the paste needs to go. Well, you need to know the centers of those for a dispenser unit. So that's a completely different file format. Yeah. But we can get most of that data, but it's just more data that's required. There's actually – there's a – Yeah.

Dave Jones: I thought the machines would come with software that would import the paste Gerber and then –

Chris Gammell: Hear the laughing, Dave?

Dave Jones: Or the paste file and then –

Chris Gammell: Dave thinks it's magical fairies that do this stuff apparently. Well, yeah. They do. They don't work too well.

Dave Jones: I was going to say it's not that – I can imagine it's not that hard to do some processing on the paste file and then determine the center point of each pad. I wouldn't have thought that would be rocket science.

Chris Church: You would think.

Dave Jones: And it would have been key to enabling a product like one of these stencil printers. Yeah.

Parker Doman: So you get that feature when you spend a lot of money on the machine. You get a reliable feature. Yeah, reliable feature.

Chris Church: They all have that feature. Yeah. It doesn't work most of the time. We actually did have a small pick and place that had a paste jet feature or paste dispensing feature. And it claimed to have that capability. Yeah. And it never once worked. No. We had to – That's incredible. We had to generate our own paste files for it. Yeah. And then the machine would just gum up all the time and put down dots that were the size of my pinky on a 0.2 millimeter pad. Yeah.

Chris Gammell: Just heat it up really slowly. Oh, goodness. Is there a lot of data that you guys need for the reflow profile then as well? Because I know that can be kind of dicey. So when we had the guys on from Salier, they were talking about the necessity of good profiles and stuff like that. Do you have to store that per part or is it just kind of general?

Parker Doman: So about 90% of the boards that we run, we really don't have to worry about it. We do check if someone's got weird parts, we do check the data sheet, make sure, hey, this can't get over a certain temperature. And so we just – and we just knock the temperature down on the oven or have to raise it up if it's like a BGA package. We've got a couple different profiles that we already know that are pretty standard for with our reflow oven. So we actually don't have to reprofile it. We just go, okay, we know that these settings give us this profile. We punch that in and go.

Chris Church: And when we're looking at oven profiles, we're really looking at like rate of error. I mean if we're talking about an extremely sensitive part, we want to know that we're not bringing that over temperature. But for most parts, a perfect profile is talking about like the difference between a 0.3 and a 0.5% error rate in the process. When we're doing a one-off, we visually inspect those. It's less expensive to verify that and fix any problems with it than it is to come up with a custom profile to avoid tombstorming for this one part.

Chris Gammell: Yeah. Great. Did you say that you read a lot – that you have to read data sheets for all these too then?

Parker Doman: Only on weird parts that I don't recognize. Like – okay. Not before seen kind of stuff. Yeah. So if someone's got an SMD connector, I'm totally going to read that data sheet and make sure it says, you can't get this part over X degrees or it'll melt.

Chris Gammell: All the plastic melts. Chocolate connectors.

Parker Doman: They actually turn – I can't remember what plastic is. It turns and looks like marshmallows. Yeah. Fluffs up. It flips up and turns brown. Just like marshmallows.

Chris Gammell: The JST, the white ones that – those like – I'll say nice. Molex connectors do that.

Chris Church: And so kind of built into that process, what we've done is we – obviously with every job, we have to kit it. And in that process, we've got in our software, the person who's kitting it has to look at all the parts he's got and he actually marks parts that bring some concern in the process for him. And then the person at that stage in the process can see those notes and say, okay, well, I need to check the temperature for this part or I need to switch to profile two for this particular job.

Chris Gammell: Yeah. That's good that you guys have those kind of things in place. I mean, is it – I mean, do your people that are doing kitting have to have a lot of expertise?

Chris Church: Generally, expertise, no. They have to be very paranoid.

Parker Doman: Always ask questions is what we tell them, right?

Chris Church: Yeah.

Parker Doman: Never assume, right?

Dave Jones: Yeah, never assume. An assumption is probably the mother of all screw-ups, right?

Chris Church: If you don't know, if you have any question, just raise your hand and say, hey, look, can you look at this for me? That solves all kinds of problems down the line.

Chris Gammell: Yep. Yep. I'm curious, too, about the style of projects coming in because I'm always curious, especially for low-quantity stuff and like Dave mentioned earlier in the show about people getting into the Kickstarter world and stuff like that, that you'd see a lot of Arduino, not clones, but derivatives, stuff like that, a lot of mega 328-type projects. Are you guys seeing a lot of spread? I mean, I'm always interested about analog, stuff like that. Are you seeing spread into a wide variety of products so far and like different parts being placed or is it kind of centralized around common themes in the maker movement?

Chris Church: Right now at this stage, you know, given that not everybody knows about us yet, what we're finding is we're attracting people who are doing either very complex things. Look at a job we're looking at right now. It's like 700 parts on a seven-inch square board to doing lots of, or a lot of times they want their BGA. I've got to get my BGA in place. I can't do this by hand. So this is a great thing to outsource for. But we also see a lot of like adapters and connectors. We have a lot of small companies using this. Yeah, yeah. And I am just shocked at how many little adapters and connectors we're producing right now.

Dave Jones: Because that used to be really hard. When I was a boy, right, before you could buy cheap, get cheap prototype PCBs made, right, it would cost you, you know, 800 bucks to get a one-off board made, right? This is what it used to cost 10, 15 years ago, right? It still costs for some people. Well, yeah. $40 connector. It was just so difficult. So you would just, you know, you'd etch the board yourself or you'd, you know, you'd jury rig it somehow. But now, you know, bam, you can just get a board from Osh Park and there's your, you know, and solder the damn thing together. And you've got a little adapter, a custom little adapter for 10 bucks or something. It's crazy.

Chris Church: And yeah, there's a lot of industrial companies actually using us. Yeah. And, you know, that's a big thing for them because a lot of stuff they're doing, they're not going to make a hundred, even a hundred of. Yeah. Yeah. Yeah. It's just a couple. Yeah. And they need it. They need it. That's all they know. And they need three of them for a customer this month. So.

Chris Gammell: When you guys are basically replacing and in, not replacing, but you're supplementing like an in-house tech. That's what companies I've been at. It's like. Pretty much. Yeah. There's a guy that was sitting there just building our boards all day because we had to, we had to have them. And he's doing protos for us. And now I can send them to you. Yeah.

Dave Jones: We would often throw this over the wall to the guy in the lab, right? We would, we would, we would have a lesson, a resident lab manager who'd take care of the parts and stocking the thing. We'd have a resident lab tech who would, you know, you could just throw this over the wall, build me one of these.

Chris Gammell: And these guys are killing jobs, Dave. Did you hear that? I heard that. I don't know. Yeah.

Chris Church: I think some of these guys are just telling their bosses, yeah, I'm just getting so much better at doing this.

Chris Gammell: Yeah. Right.

Dave Jones: Yeah.

Chris Gammell: You guys should have shell corporations so that it's like, you know, like it's a different name on the credit card statement. So it's like, you know, something, something training or something like that, you know,

Chris Church: and then the board show up. Parker's been known to, you know, take cash out on the, out on the, uh, the driveway out there. Buddy, you want to board a sample? Want to buy a board?

Chris Gammell: Got that for you, man.

Parker Doman: I got some breakout boards at a great price.

Chris Gammell: So I'm curious about, I mean, especially with this, this adapter board thing as well. I mean, like, so pulling in, uh, design files for, for connectors and stuff like that. Is that, is that a big, uh, a lot of, uh, manual side for you guys for like, uh, you know, new, new connectors and stuff like that? Is it, is that normally going to be hand placed and everything like that?

Chris Church: Yeah. The, the, there's a lot of these projects for the volume we're looking at where, you know, we get like two or three in an order and, you know, it's something you can snap in and half a second. We would do those by hand. Gotcha. And we do the majority of those by hand. Cause it's just, again, it's so much easier and faster to do it that way.

Parker Doman: Yeah. And it's only a couple. It's just, it's just easier to say, Hey, put this part there and done. Yeah.

Chris Gammell: Mm-hmm. Yeah. Yeah.

Dave Jones: All right. I, I seen mostly, this ties into Chris's question about the, uh, Kickstarter and all that sort of, you know, the type of boards. Um, are they mostly like just designed to be like bare board products or are they, or can you tell that there are, they've carefully designed this board. It's going to go into a custom housing, blah, blah, blah, blah, blah.

Chris Church: The vast majority we see are bare boards. Um, I would have thought so.

Dave Jones: Yeah.

Chris Church: But we are trying to encourage the customers to make them look better at enclosures. You know, you can, you can make a little more money if it looked like a more professional

Dave Jones: product, you know, and coaching and branch off there for we'll, we'll do your custom enclosure for you. You'll get into 3d CAD side of things.

Chris Church: We'll take your 3d CAD file and I will say we want to help our customers solve the enclosure problem this year. Oh, nice. Okay. That's soon.

Chris Gammell: Wow.

Dave Jones: Excellent. Excellent. And go to buy one of those, you know, a hundred thousand dollars 3d printers and yep.

Chris Church: Nobody buys those anymore though. They just lease them out.

Parker Doman: So just, or, or buy $300, you know, $1,000 printers instead. That's what I want to say. You'll see that a $200 one that's on Kickstarter now, $179 Delta bot. No. Really? It won't happen. It won't happen. They cannot make a profit on that. They will fall. They cleared their Kickstarter goal in one day. Oh God.

Dave Jones: Of course they did, but, but they'll fall. They won't be able to produce it. I guarantee you.

Chris Church: Well, even if they can, I mean, can they support the, uh, all the service calls, you know, all the support. Yeah.

Dave Jones: You know, they won't be able to, somebody did an article on this. It's impossible to make a 3d printer under 300 bucks. You can't do it.

Chris Church: Yeah.

Dave Jones: I think I saw that one. It's just too involved. It's just, there's too mechanical, too many mechanical bits to it. There's a limit, you know, it's not like boards and components, you know, you can churn out, you know, PCBs ever cheaper, you know, um, by using jelly bean gray market parts and then getting your labor in China, blah, blah. But yeah, complex machines, um, take money to build. Unfortunately.

Chris Church: People often forget the, the warranty, the returns on those, you know, you're going to replace, you know, especially on a 3d printer, you may replace that. I wouldn't be surprised if someone was replacing one out of every 20 or one out of every 30.

Parker Doman: Yeah. With a complicated machine like that.

Chris Gammell: Yeah. Yeah. Alignments and tolerances and all that crap. Right.

Parker Doman: Even like complicated things like car transmission. So I was looking at like 2% is a good number for failure. Yeah. On a transmission. I'm like, that seems a little high to me. It does.

Chris Gammell: Especially if you're driving in the car on the highway at the time. Yeah.

Dave Jones: Now, what about, what about the test side of things designed for test? Um, do you build custom test jigs for people or do you just want to stick to the assembly?

Chris Church: You know, we, we can't actually, for all of our, we, we have a handful of customers. We actually do end to end Kickstarter services for where we not only source everything, we manufacture it, we program it, we test it, assemble and fulfill it for them. So in that process, we got to make sure that that's a pristine product going out the door. Otherwise they look bad. And then we, by extension, look bad and don't get any further business. Um, so generally we will, uh, either the customer will actually design the test jig for us or will, they'll pay us to design the custom jig for them.

Parker Doman: And we do help them if they're still in the design phase to make sure, Hey, we can test this and it won't cost an arm and leg.

Chris Church: Hmm.

Dave Jones: If you add these pins on here and this, if you add a test port on there, that'll like,

Parker Doman: you know, exactly. Yeah. Or design the panel for testing so we can test, you know, all 18 at once and stuff like that.

Dave Jones: Once. Yep. Awesome. It's. You mentioned the fulfillment. How much of a pain in the ass is fulfillment? How do you go about doing that? Do you do it yourself or do you, as you like, just sit in a room, right? We're going to do it all night.

Parker Doman: I'm hacking the ship and with the right now it's, I actually do all the shipping.

Chris Church: That's my night job. And I, and I write all the software. Congratulations. He runs into while he's doing that.

Parker Doman: Yeah. It's actually not too bad right now. Um, we got a lot of good software that makes my, makes that a lot easier.

Chris Gammell: Wait, is this what you meant Parker? When you said you moved closer, does that just mean you're sleeping in the shipping room department?

Dave Jones: Yeah. Yeah.

Parker Doman: Yeah. Sometimes. Um, I don't know. We were, we were actually one, one, uh, warehouse down. Um, I had a cot in there.

Speaker ?: Yeah.

Parker Doman: Stardust.

Chris Church: He also, he also mentioned that there's a brewery right next door. They were literally right next door then. So that's part of the reason why I was there. Leaving the dream. Yeah.

Chris Gammell: Why are all their labels on upside down? I don't know. Wreck me up in an hour. Yeah. Right. Yeah.

Chris Church: So that, that, that's, that's really a thing. It's, it's actually, there's not a lot of work to do in it until something goes wrong. When something goes wrong, that's when, that's when all the work really, you know, kicks in on that. Dealing with the postal service, dealing with the customers. And by this, I mean the end customer, the people getting the product. You know, nobody. Yeah. It doesn't, you know, nobody's happy when they don't get the thing they're looking for. So just trying to, you know, make, make, keep everybody happy in that process. And so that's, we tend to work with the customers as well. Our customers to, you know, ensure that they're, they understand that there's going to be some problems and we produce a little overage so we can make that easier on their customer. And just, you know, no questions asked. The package is lost. Let's send out a new one. Like right away. Yep. Yep. Always cheaper. Yep.

Chris Gammell: Yeah.

Dave Jones: Have you ever sat there and thought, geez, I think we're biting off more than we can chew, trying to do too many things, trying to do everything.

Chris Church: Oh man.

Dave Jones: Has that ever crossed your mind or, or has that been your goal and that's what you're going to do regardless?

Parker Doman: It's been a goal, but I think every startup feels like that sometimes. Yeah, right.

Chris Church: It's a good night when I don't go to bed worried.

Chris Gammell: Yeah.

Chris Church: That's good. Yeah. Usually means I had maybe one, one glass of wine too many or something, but.

Speaker ?: Yeah.

Dave Jones: Can you see that happening down the line that you're going, well, you know, designing cases or testing and shipping is just not the thing we want to do?

Chris Church: You know, we do actually do get that question a lot, mostly from potential investors. Like, don't you think this is just too much? Right. Don't you think that there's not a way you can make more money off these people by doing less for them?

Dave Jones: Which is all they care about, right? Making more money out of the people. They don't care about your philosophy and goal. Right. So that. Like, oh, they might have a token. They might care a token amount, but, you know.

Chris Church: So that, to be honest, you know, it always seems like it's too much when we had a problem today, we're behind or, you know, something's broken. But I just, you know, if we weren't doing all this, I don't know that I'd want to do this. You know, I don't want to do just PCB assembly. That's kind of boring. Right. I feel the same way. Yeah, yeah, yeah.

Parker Doman: It's, it's, it's, I like to build people's products, not their boards. Right.

Dave Jones: Got it. Yeah. I like that. I can, I can dig that. Yeah.

Chris Gammell: I have to circle back around to my question about me being sloppy. And so if I send you guys a bomb with like a really bad bomb, and I think you answered that some people, some, you know, you said like 70% of the people are in decent shape. But like, my question there was actually more around like, so if I send you a file that all of the resistors just say 1K 0805, like how, how do you?

Dave Jones: Well, first of all, they spend 10 minutes just sitting around cacking themselves laughing. Chris Gammell, what a dickhead. I mean, no, because I think, I mean like. What he did? Can you believe this?

Chris Gammell: I mean, I don't see the problem. I mean, like, honestly, I don't see the problem there. Like, if you're going to send it to a contract manufacturer, they're going to replace them all anyways, right? I mean, they're going to, they're going to go and take house parts at any CM. So, I mean, can you guys handle that?

Chris Church: So if you just had a board full of 1Ks and you actually meant for those to be 1Ks, yeah, that's not a problem for us. We wouldn't be happy with you because we might lose some money on that deal. Oh, of course, yeah, it's better to use your part number. Right. Because if it's all the same resistor as the house part, you know, we may not actually make any money on that board. That's not a suggestion. No, no.

Chris Gammell: So what I'm actually asking about is, is like the detection algorithm that you guys have. So if it says 1K0805, does your algorithm then go and say, here's our house part, and then it'll just recommend that for all those parts? Yes, correct.

Chris Church: And even worse, because it's trying to simplify your bomb view, or worse, depending on if this was a mistake on your part or if this was intentional, it'll actually group those all up automatically into one bomb line as well.

Chris Gammell: Yeah. No, that's, I mean, that's what I'd expect. I mean, even if it was like, not that they would all be listed separately, although Kaiky cannot put it like that sometimes, which is weird. But yeah, no, I actually, I just meant like just the detection side of things. Oh, yeah, yeah. Because, you know, like if I'm sending it off to a CM, I hate, I hate the passive stuff. I mean, like if it's just the jelly bean, I think that's, you know, that's nice to be able to just have it loosey-goosey, 1K, 0.5, 1%, you know, I don't care.

Chris Church: And that's like the thing I was talking about earlier. If you had a 1K and a 995 and a 1,100, our system would automatically try to suggest the 1K for all of those lines to you.

Dave Jones: Hmm. Right.

Chris Gammell: Is there a way to pull that back out into the files as well? I mean, back into...

Chris Church: Oh, to get your bomb back? Yeah, like back annotate almost? Not yet, but it's coming.

Chris Gammell: Okay.

Chris Church: Yeah, this is going to be an export function soon-ish. Because we generated actually a lot of files from your project, everything from SVGs to our format data files and some things your CAD might not normally produce, like a paste file or a, you know, a combined placement and build the materials. And we want to give you the ability to actually get those back out so you can use them for your purposes.

Chris Gammell: One of the things I like the best about this system, it allows the bomb, no, sorry, the schematic as a design, as the control document. That's what I actually love about it. I love that about all these kinds of systems where if I want to change something, it should be, you know, I go submit it to GitHub with that one change, that 1K becomes 2K, and then I don't generate a spreadsheet. I don't have the decoupled from the schematic. You know, like I want the schematic to be the be-all, end-all truth of my world. And I think that you guys allow that.

Chris Church: And you and I think alike. I wish more customers thought that way.

Chris Gammell: Yeah. Well, I've screwed up enough spreadsheets and disastrous effects that that's the problem, right? I mean, and I think, I mean, that's one reason I'm such a big fan of KiCad, too, is that, you know, one of the problems, one of the reasons that I've always said that that's impeding schematics as a control document is, you know, if you have to pay five grand for each seat of the schematic tool or the, you know, the entire layout tool, no one else can go in and change it, right? You know, it's only going to be that one person with the seat. And that sucks. Now, if it's open source or if it's really low cost, then anyone can go in and change it. You may not want them to, but that's possible. And you could have a production person go and change that file and, you know, revision control it and get approval and stuff like that. But you can actually change it at the schematic level. And that's really important for me.

Chris Church: And I 100% agree. I mean, we really, you know, just coming from an open source background, I mean, we really need to look at that original source code as sort of the, what's the word I'm looking for here, the exact representation of that, not all the artifacts you produce from that. Right. Yep. The canonical representation of it. There we go.

Chris Gammell: Is that on the bingo list? We don't talk software enough to say canonical, so.

Chris Church: But, you know, I mean, from a perspective, I come from a software background. So, you know, to me, it seems odd that there's something other than the source code, which defines the, you know, how to actually build this product. Right. And that has always seemed weird for me in the hardware world. And, you know, I've always worked with our customers to try to say, let's do something other than an Excel file. Yeah. But let's not use an Excel file. So, yeah.

Chris Gammell: Can you guys pull off GitHub or? Can we? No, no.

Chris Church: Not right now. We probably will later. We haven't actually had a lot of demand for that because most of our customers actually are using private or proprietary designs.

Chris Gammell: Oh, yeah. But there's private repos. I mean, you could. But it's, I think, just a lot of hardware people aren't using GitHub or just revision control in general. I know that, I mean, Dave, you said Altium has revision control, right?

Dave Jones: Yeah, it does. It has its own flavor. Yep. Yeah.

Parker Doman: I wish more hardware people would use something like GitHub. Right. Because once, because I'd never used revision control like that until I started working with Chris. And we used SVG. SVN. SVN. Subversion, yeah. Yeah. And that was, that was okay. And then we started using Git and I really like it.

Chris Gammell: Yeah. Yeah. I mean, just from, and especially if you're using, you know, text-based CAD files as well. So you can, I mean, you're never going to see a perfect diff, but, you know, if you can start to diff files and actually see some legit differences on, you know, especially BOMs or even just the source file, the schematic itself, seeing new, new entries and stuff like that, that can really help with that kind of thing.

Chris Church: And we've even seen, you know, from GitHub, I think they've, they've done, especially for some of the 3D models, I think they have done visual diffs on the 3D models. Yeah. And I think if more people were using, you know, open source, you know, EDA tools with that, we'd probably see visual diffs from them on the schematics and the board designs as well. Visual diff Gerbers. Yeah.

Chris Gammell: Yeah. Yeah. There's a guy that did a visual diff on, with the Gerber generation script and then you can, he did an overlay and then, um, yeah.

Chris Church: Yeah. I saw that actually, uh, I used that for a prototype we built. Oh, nice. Earlier, or earlier last year, we actually had a feature in the UI to do diffs and it was really cool, but it was just way too much to deal with. So we took it out.

Chris Gammell: Makes sense.

Dave Jones: On the software side of things, how much work has gone into producing all this magic? A lot. All this software magic. I was going to say it's, it's probably the majority of your development, right?

Chris Church: It has to be. Yeah. Uh, uh, the, the UI actually, you know, we, we kind of take a both, uh, we, we, we take sort of an API first approach. So we start with an API for any capability and then we put the UI on top of that. It's just so much easier to work from that, that perspective. It allows us to, we do what we call a microservice architecture in the backend. So we're able to split up the workload amongst many different servers. So the actual, the user interface part is, is, is not a whole lot of time. It's all of those models and controllers and everything behind it that really take the most time.

Dave Jones: Um, and how many people you got working on all this?

Chris Church: Uh, right now we have two people developing full time. Um, okay. We're actually adding two more here as soon as we can find the people. Yeah. And Chris is actually one of those two people. Right. Right. Yeah.

Dave Jones: So we actually only have one other guy. This is what we specialize here on the Amp Hour, finding people. Right.

Chris Gammell: We, we can always, uh, put out the call at the end here.

Dave Jones: We've had, what, three or four guests at least who have found people through this show.

Chris Church: Yeah. Yeah. Hey, we're hiring and we pay industry, uh, standard salaries too for a startup.

Chris Gammell: Whoa, guys. Really sell it on that one. Yeah. If you look at a bell curve, we are direct center, guys. Absolutely.

Chris Church: Normally on the startup, it's way down below. I know. Right, right. Right. Yeah.

Chris Gammell: Yeah. At the bottom with equity. Right. Yeah.

Dave Jones: Well, is it, is it like a stock thing or is it like just a regular job?

Chris Church: Uh, all, all of our, all of our employees, uh, get, get stock and they don't get options. They actually get real stock. Common stock. Right. Okay. Very cool.

Dave Jones: So there you go. So you're looking for programmers.

Chris Church: Absolutely.

Dave Jones: There you go. We're clever programmers out there. Um, if you want a job in Houston. Yeah. Home of the big guns.

Chris Church: They can be remote too. Oh yeah.

Dave Jones: Ah, yes. Right. There you go. Excellent. We'll put details down below where people can sign up and we, we should start charging a commission, Chris.

Chris Gammell: No, come on. These guys, they offer their time. I mean, come on, man. Yeah, I know. Yeah.

Dave Jones: Okay. There I am. Another fraybie.

Chris Gammell: Another fraybie.

Chris Church: We'll give really big discounts to anyone who helps us hire. There you go. Yeah. Coupon codes for, uh, for boards, right? Yeah.

Parker Doman: Chris Gammell's secret super coupon code. There you go.

Chris Gammell: Uh, speaking, speaking of coupon codes, do you guys, I, the last, this is the last thing I wanted to ask. I think Dave, Dave, we should, we should probably wrap this up eventually, you know, hour and a half so far. Um, are you guys seeing, I mean, other than just the normal breaks, I always wonder like if CMs, I mean, you guys are CM ish, right? I wouldn't say just, just CM, but, um, are you guys starting to be able to bundle enough or do you plan to be able to bundle enough to start negotiating like larger scale discounts with the, with the big distributors and manufacturers?

Chris Church: Yeah, that is definitely on our list. We're not there yet, but we're getting there pretty quickly.

Chris Gammell: So I'm always amazed. And I think we've talked about on the show before, it's just like, like to do so, you're still going to need to talk to a salesperson, like for all of your code and all of your brilliance and, you know, like all of this front end, backend API stuff, like you're still talking to Jimbo, the sales guy.

Parker Doman: That's how the house parts are so, um, inexpensive is, is we buy them in bulk. So we, we give the, the, the discount we were able to get, we pass that on.

Chris Church: Yeah. And, you know, we, we do, we have been outside of just the distributors. We've been working with the manufacturers too, to try to get the, you know, a better discount, you know, get them to get the discount for us through their distributors since they're ultimately the source of this and they have a lot more sway than us. And we're, we're finding that they are very amenable, amenable to that when we're providing value to their customers. So we usually, we have to give a little take a little bit.

Chris Gammell: Yeah. I'm always surprised that like, like, so like when I knew the, I mean, I've kind of fallen out of favor with the folks in this area because I don't really make anything around here anymore. But like they always said their best customers were like the, the sales guys I knew who were earning commissions were the guys who were in the manufacturers in town, you know, the, the manufacturer salespeople and the FEs and stuff like that. They, they got a big chunk of their money from the actual manufacturing on the ground in Cleveland because, you know, it's just like direct and they can tie it directly to that, that facility. So yeah, salespeople probably love, love and or will love you guys.

Chris Church: Yeah. And I mean, that's, that's a part of it. And it's too, it's like we're part of our house part, really our house part model provides a lot of a benefit back to the, to their users as well, the manufacturer's users. And we found in some cases, they're really keen on the idea if they're commonly using a component that's very hard to find, uh, we suddenly get a lot of help from them when we agree that we'll carry that as a stock house part. Uh-huh. Yep. It's your bargaining chip. Suddenly they're like, you know what? Hey, if you guys are willing to carry that, we'll, we'll, we'll help you get a discount on this part from this distributor over here. So.

Dave Jones: They love the ability to schedule things. They just love it.

Chris Gammell: Yep. Exactly. Gets more excited. So, so people, if they go sign up right now for the service, I mean, it's just pretty much upload a bomb. They can go right away almost, right? I mean, like 15 days later.

Chris Church: Go through your design files and, uh, you know, take a look through your bill of materials, make sure everything's correct and hit order and go over there. It is in beta. So. There'll be bugs. There'll be bugs. But if it's a pricing bug in your favor, well, that's the price you're paying. Like an apple. I'm going to come back. Thank you. In your favor.

Dave Jones: Well, thank you very much, guys. It's been awesome.

Chris Church: Hey, no problem. Thanks for having us. I've really enjoyed this.

Chris Gammell: Oh, yeah. We would love to talk to you guys again in the future, especially as you, you know, continue. I'm sure, like, come on, like this is going to do awesome. I'm sure it's already doing awesome. But yeah, this is, this is, this is a big deal, folks. If you didn't know, this is a big deal. Take it from me and Dave.

Dave Jones: Do you have any direct competitors?

Chris Church: Right now, I mean, the closest we're going to get is going to be like CircuitHub and SnapEDA. Right. Again, for that, that experience on the upfront side. But we haven't seen anyone meet, meet our one-off price yet.

Parker Doman: I have not found out either. Or, or has like a front end like we do and actually run the line.

Chris Church: Yeah. At the same time.

Parker Doman: At the same time.

Chris Church: But we do know that's, that's coming, by the way. I mean, you can look at, look at the market out there. The EMS companies know that they're behind right now on that, that user onboarding friction phase of the process.

Parker Doman: Another bingo word, friction.

Chris Church: And onboarding. Don't forget that. Yeah, onboarding.

Dave Jones: Oh, yes, yes.

Chris Church: I was talking to my board today, so it's kind of, but they know that they have these problems getting new customers in the system and they see us, they see CircuitHub, they see SnapEDA. And I think we're probably, I took a look out there, we've probably got three or four more coming like that, which is like an instant easy quote front end. And the manufacturers, the contract manufacturers see this and you, I can guarantee you, they're going to be building those interfaces over the next year or two here. You're going to start seeing that from other manufacturers like us.

Chris Gammell: Bully for us, Dave. I know.

Dave Jones: Fantastic for us. You guys might go out of business, but yay for us. Oh, no.

Chris Gammell: No, they'll stay ahead of us. I'm sure, yeah. Yeah, yeah. All you guys, they're all doing great. You know, like it's, thank you for pushing them. You know, that's the main thing. They could get away with it for so long because it's like, oh, single board, you're not worth our time. And now it's gotten to the point where it's just, you know, there's a market for it. That's great. Yeah. Now there's software. Awesome. Yeah.

Chris Church: And they may be still see the single board as not worth it, but that whole like, you know, that instant upload your files, we'll quote your parts and everything. That's coming. That's coming from everybody here soon.

Chris Gammell: Gotcha. Very cool.

Chris Church: Good stuff. Thank you very much, guys. Hey, no problem. Thanks for having us. Yeah, thanks.

Chris Gammell: Oh, one last thing. Sorry. One last thing. Where do people send their software resumes and or hardware resumes if Parker's hiring as well?

Chris Church: Oh, to jobs at macrofab.net.

Chris Gammell: All right. All right. I'll post a text link to that.

Dave Jones: Why is it .net and not .com?

Chris Church: It was taken and they won't sell it. Oh.

Dave Jones: Bastards. Is it a squatter or is it a real user? It's a squatter. Did you ever think, well, we'll just change our name because .com is everything?

Chris Church: No, it's not, Dave. It is.

Chris Gammell: It used to be.

Chris Church: No, we looked at it and this is the name we wanted. And eventually we'll get that.

Parker Doman: I wonder if we open up another shop somewhere, we're going to be macrofabs.

Chris Church: Oh.

Parker Doman: Great register.com now.

Chris Gammell: Yeah. All right, guys. Thanks again. All right. We look forward to trying out the service. Thanks, guys.

Dave Jones: See ya.

Chris Gammell: See ya.

Speaker ?: x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x