Embedded - 407: Boards Are Like Sandwiches

Episode Date: March 24, 2022

Mihir Shah of Royal Circuits joined us to talk about how PCBs are fabricated and how companies are funded. Mihir was CEO of InspectAR before they were acquired by Cadence. Mihir works for Royal Circuits and runs a newsletter called TheAnalog.io We talked about InspectAR on Embedded 384: What Is a Board File? with Liam Cadigan. Transcript for this show This episode is sponsored by Newark, a leading international distributor of industrial and electronic components. From design and testing to production and maintenance, discover why so many choose to partner with Newark!

Welcome to Embedded. I am Alicia White, alongside Christopher White. This week, we're going to talk about circuit boards and venture projects. I'd like to welcome Mihir Shah to the show. Hi, Mihir. How are you doing? Hey, guys. Thanks for having me. Could you tell us about yourself as if we met at Supercon? That's a good, that's good context because I was there at Supercon and that's one of my favorite events to go to.

So I'm an electrical engineer by trade. I started my career at Tesla as a hardware engineer and then I worked at Taser on some of those actual Taser weapons and pretty interesting hardware development there. And then I went in after a series of different stints working for Royal Circuits and this and that, we actually started a company called Inspect AR. And you had one of my co-founders on the show, Liam Cadigan, and that company was awesome. We built it, grew it, and it was acquired by Cadence, but geez, almost two years ago now. And since then, I've come back to Royal Circuits and I'm really leading up more of the ventures and growth part of the business as we continue to scale in terms of PCU manufacturing and then all the startups

in terms of software and hardware that are kind of tangential to that business. Excellent. Well, we have more questions about that, but first we're going to do a lightning round. Are you ready? I'm ready. Optimal number of founders in a startup? Three. Favorite flux.

I don't have one in particular, but flux is good. More. Best PCB color. Purple. How many funding rounds is too many funding rounds? Hard to say. It depends on the business. But let's just say five.

Five. That's it. That's when I quit companies. Worst way to pitch an idea. Isn't that in the bathroom? You have to let him answer. That's true. True.

I mean, I've heard it's been done before, but how do I say it? Not confident or lacking confidence. Favorite kind of aircraft that you can fly? Me, Cessna 172 Skyhawk. Because that's the only airplane that I really can fly now that I'm a private pilot as of yesterday. Congratulations.

Congratulations. IFR or VFR? VFR. Of course. really can't fly now that i'm a private pilot as of yesterday congratulations congratulations ifr or vfr vfr of course there's no ifr yet you can't go for both at the same time instruments next instruments next complete one project or start a dozen oh start a dozen do you have a favorite fictional robot? No. Ooh! These lightning rounds are good. I'm like, oh man, I'm wracking my brain.

And now, a word from our sponsor. Why choose Newark for electronic components? From humble beginnings as a small electronics shop in 1934, Newark has grown into a leading international distributor of electronics and industrial components. They offer products, services, and solutions that support customers in a wide variety of applications. Industrial manufacturing and maintenance, medical product design, automotive, avionics, telecommunications, security, and more. From design and testing to production and maintenance, discover why so many choose to partner with Newark. Newark has operations in the U.S., Canada, and Mexico. They are serviced from a regional distribution hub in South Carolina. New., Canada, and Mexico. They are serviced from a regional distribution hub in South Carolina.

Newark is committed to supporting local language, currency, product, and shipping needs across North America and around the world. As part of a global business, Newark has access to stock and strong relationships with electronic manufacturers. They can easily serve all of your industrial electronic needs. Learn more at Newark.com.

That's N-E-W-A-R-K dot com. And we have a coupon for those of you in the U.S. Pod Save 20 will get you 20% off. That's P-O-D-S-A-V-E 2-0. Thank you to Newark for sponsoring this show. You mentioned Royal Circuit Solutions. Yes. Could you tell us more about that? Yes. Royal Circuits is a company that my dad really built and grew from effectively the ground up about 15 years ago after he had started a company called Advanced Assembly. Royal Circuits as a group effectively is one of the fastest and largest quick turn prototype printed circuit board manufacturers in the United States today. So huge variance in customers. And the idea is ultra quick turn,

one to five days and increasing complexity, right? HDI boards, multi-lamp cycle, this and that for a broad range of sectors, everything from automotive to aerospace, to medical, to consumer. So we manufacture circuit boards here in the United States at the prototype level really, really fast. Do you do assembly too? Yeah. So we do assembly with one of the companies in our group, Advanced Assembly. And then we also work with a few other kind of more mid-volume CMs, but we do prototype assembly as well, full turnkey. What's the hardest part of making PCBs? The hardest part of making PCBs? Well, rigid or flex. You know, the hardest part is probably, as far as the factory goes, is making sure you have the right people and making sure that they are trained correctly. Because,

yeah, there's a lot, a lot, a lot more labor involved than people think. And it is actually highly skilled labor, like running, operating a drill machine, laser routing, things like that. It's not trivial. That's probably the hardest part of running these businesses. What's actually, I'm pretty ignorant about PCBs other than, you know, you do your CAD stuff and then you ship it off to a company like yours and then magically in a couple of days or a week, things come.

What's actually involved in the process in terms of machinery and stuff? Like, how is that actually done? And how automated is it? Everyone's going to say, oh, there's varying levels of automation. But at the end of the day, there are a lot of hands that touch any sort of board, even the most what we would call simple, like a basic, basic two-layer through-hole only, you know, real simple board. Even that one is a ton of labor, which is then when you start doing the math of how these boards from China are so, so cheap,

you start to say, well, this, how is that even possible if this many people are involved in the manufacturing? It's automated, as they say. So you still have to go and move the, you know, you have to collect the material. You got to image it. These are large, pretty expensive machines

that actually have to do pretty detailed lithography onto the panel itself. You got to develop the resist. You got to etch it out. So it goes through a pretty heavy chemical process in that regard. We got to move it between different rooms and departments and buildings. Then you have to go and actually, if it's an inner layer, you got to go check them, making sure it was etched correctly. You don't have inadvertent shorts or even opens that you didn't expect after the etching process, after which you may have to either scrap it or re-image it or re-etch it. And so there's a lot of fine tuning depending on the copper density and the actual fine lines of the traces, the thickness of the trace in space.

So there's all these different variants, but you have the, how should I say, the machines for the most part are the machines you have on the floor. And that's one of the things that we do very, I guess, differently or we're pretty aggressive about relative to our competitors is our investment in actual capital expenditure is probably like 3x the average shop certainly of our size meaning we have two of every machine at least and we're always updating and upgrading and buying new machines uh latest and greatest from all over the world because you have to do that to maintain so anyways to answer your question i just got

through like the first 10% of it. You know, I think I go through solder mask and then you got to go through different levels of drill. And if it's a multi-layer board, four or six layers, and you have internal drills, like a buried or blind, now you have to go, you know, there's separate travelers. The board is in three different places at once. Effectively, it's going through the drill area two, three times, going to be laminated a few times, and that just will take literal time. So now what could have been maybe a one-day turn is that fast to say two or three days, just because you need four to six hours on the press. And then you have to go press the whole thing together afterwards. Add it up, takes time. But there's a lot of things that go into a board that at the end

they might just be lighting up a blinky or be controlling avionics for a for a rocket ship so there's a lot okay now you've said you've only gotten through a piece of it and yet i need to go back to step one and ask you to repeat that as though I were a five-year-old. Yeah. You have a board. I assume you start on a middle layer, or do you start on the bottom layer or the top layer? Boards are like sandwiches that are built from the inside out. So you start from the core, which is the center of the board.

So if it's a four-layer board, you're going to start with a core that has copper on either side so it's going to be layers two and three you're going to image those whatever was in the design file you basically blast the light over the resist in the areas where you want to have or not have a copper on so you put you have you have a something a piece of fiberglass yes and then you put copper on both sides and then you put on this resist that will mean that the copper stays there as you do something to get the rest of the copper off exactly yeah you put this resist on and then you image it so wherever the light touches that part that is the area that the resist is going to harden so that later on when you put it through uh the developer the the the resist that wasn't hardened will be washed away leaving you

exposed copper that you're then going to go and etch away so it's a negative process in that regard because afterwards once you get rid of the hardened, you're just left with copper in the areas that you want it. So you can effectively have point-to-point connections. And it's not just one big short, which is what you started with, one big sheet of copper. And you said developer. This is like photography was. Yeah. It's the same process.

And so there's like a chemical bath that makes the chemicals do things well do you remember radio shack you could buy bare pcbs and you could paint on resist oh i have some and then you could buy a bottle of etching which they wouldn't give to you if you were under 18 and then you could experience. This sounds like automating and building that process up into something more sophisticated than somebody pouring acid onto a board and hoping they didn't spill it. Yeah.

That's a basic process, right? That's about it. Yeah, it's not something that's... I guess the processes themselves are not overly complex. Once you familiarize yourselves with them, it's more the things that are just straight up manufacturing right what happens if your photo resistor your developer goes down well you could use the other developer that we have that is used for solder mask it's basically the same chemistry except now you're kind of mixing some of the stuff so the

solution gets a little contaminated if you left that going for a while, it may have adverse kind of effects on the board. So it's just thinking about how do you manage a factory, making sure you have two up of everything. You got the parts to actually fix things on the fly because a lot of these machines are not from the US. And so the people that service them are not, even if they wanted to, able to come here in a day's notice oftentimes.

And with COVID, it may take them two, three plus weeks to get over here. You can't be down that long if you're a quick turn shop. You can't be down for like six hours. So having the ability to be self-sufficient and work and manage a lot of the machinery on your own is another interesting part about the business where you kind of become you run your own machine shop as well and kind of a maintenance crew always running yeah i have a question it just occurred to me and it's sort of off topic but you mentioned chemistry um and copper is expensive so you etch away most of the copper that's on these boards to leave the traces yeah can you recover the copper from the solution afterward or is it gone it's gone and it's not like huge amounts of copper for it's not huge amounts the copper is actually pretty thin on the boards it depends of course

half ounce one ounce copper over that uh standard 18 by 24 sheet which is the standard panel size that we run through the shop. Okay. But yeah, it's etched away. Okay, so we have a board that has a core that is non-conductive. We have two layers of copper, one on each side.

Let's say we only want a two-layer board. Yeah. At this point, you put on plastic stuff the resist no no no after the copper the solder mask

are you talking about the solder mask well you can't put the solder mask next you have to put the glossy thing and then the silk screen on the glossy thing

what what makes the what makes the board a certain color that's the solder mask no you got it right oh it was the solder mask okay yeah those names are they're very similar and if you haven't heard them or you don't hear them frequently they're easy to interchangeable in that regard but solder

mask is the color and it could be any color right royal we have customers that want pink solder mask matte black uh taser we would do different color solder masks depending on the project sometimes a revision just so visually you're like okay that is a rev two board that's a rev three don't mix them up because they otherwise look pretty similar i used to do that yeah every rev got a different color and it was and you tried rainbow if you could yeah we've done tie dye uh we mixed the colors i was gonna say can you do plaid yeah uh there's there's a there's a lot of cool uh yeah there's a lot of fun colors that we've done i thought the solder mask was the thing that you put on after that like had all of the designatorsators? Not the designators, but the thing that you then, you put down, you put on, you glop on some liquid solder and then you wipe it off and then you pull up that piece.

I thought that piece was the solder mask. Oh, that's a stencil. The stencil, yes. Could be a stencil. Yeah, that's after the fact. Or you're talking about the surface finish, kind of like doing a hassle finish or an eneg if you want gold uh finish on top of your pads which is very common today most most customers uh choose that why uh you just better conductivity thermal properties it doesn't tarnish really in the same way that hassle may um some aerospace

people do hassle the the one with lead because the hassle is lead free but some people need that for extra reliability some aerospace applications but for the most part by and large uh we we see enig is the most popular surface finish okay and then okay so you have the solder mask which controls the color and then you have this silk screen which is kind of like you just put it on a t-shirt. Yeah. And it's just like that. The process looks basically just like that. We have printers and things that will automate it and cure it, but sometimes actually you get really, really high resolution. It actually is better to do it manually.

Very similar to how you do it on t-shirts. You just put it down and you squeegee it forward and back. That is kind of manual. Yes. And you guys are in Northern California. So really, I invite you guys to come to our factory anytime. We'd love to give you guys a tour. And honestly, this is extended out to anyone listening.

If you reach out and you're in the area and you'd like to visit one of our facilities, we'd be more than happy to accommodate anyone and give you a tour. Super proud of what we have and how we built the company. You're in Hollister, right? Sort of near Gilroy?

Yeah. Yeah, the town right after, just south of Gilroy. So 30, 40 minutes south of San Jose. Okay, so now we have our PCB and I've assembled them. So I assume that's how you would assemble them at your location yeah we well we got some fancier larger machines like pick and place machines but

and and assembly is not a lost art still and it's I think something that if you can do it it's pretty awesome yes and you your tools are far better than mine. Between pick and place and not having to use a hot plate to do that soldering. It'll save you a lot of time. Especially with a lot of components or packages that are more leadless. Pretty hard to get down and make sure you got a good connection. You may want to go with a professional assembly house. So having worked at Tesla and Taser, did that give you experience with different PCB houses?

Oh yeah. Oh, and we visited so many. And so it was a great experience being on the customer side and actually having to remove my bias and be like, well, one's my family business, but hey, the guys at Tesla, for a number of reasons, oftentimes had no technical basis behind it. Any of these companies, not just picking on Tesla, I'm just saying in general. For most boards, if they're not overly complex or they require a very specific configuration or large audits most shops or

there's a group of shops that can build on i can give you an interesting tidbit on the history of this industry or kind of how the makeup is today but you let me know if that's interesting to the go ahead yes listeners what you'll find is on certainly on the board side, about 10, 15 years ago, there was like over 2000 board shops in the United States. Today, in terms of fabrication, bare boards. Today, there's like less than 150, of which really 15, maybe 20, not even, make up like 90% of the market here for prototype quick-turn fabrication because the industry is pretty fragmented. And what you see is most of these shops are the 2000, 1000 plus shops. They didn't really invest in the equipment and upgrade beyond what was really the majority of the business 15

years ago, two layer, maybe four layer, simple surface finishes through hole for the most part like real easy boards and now you got to be able to do high layer counts really really dense um you know imaging fine trace and space i mean you're getting down to two and two that's two mils the human hair is about three mils so imagine a trace like a wire on a board being two mils thick less than a human hair is about three mils. So imagine a trace like a wire on a board being two mils thick, less than a human hair with a two mil space in between. So being able to do that just requires expensive equipment. Like we just bought a few new drills. Each of them was over $350,000.

And these are pieces of equipment that we're buying nonstop. Every few months, you'll see new equipment come to the shop. So most shops weren't able to keep up with the demand and they became brokers or just went out of business. The shops that are remaining, you're seeing now a pretty heavy consolidation amongst a few of the larger shops and then also private equity firms coming in

because these are incredible businesses, great margins, but largely owner-operated from, I guess you could say, a previous era. So it makes sense that you need a lot more money to keep these going, and there's value in putting them together. So what you're seeing now is a lot of consolidation in the market with these more financial buyers coming in and helping growth so the customers can continue getting what they need, especially on the more high-end side with aerospace, military defense, things like that. So the industry is undergoing a shift, but the customers are only being better served because there's more money coming into the space. Why did so much of it go offshore if it's a high-margin business? Well, so the high margin stuff is the really complex stuff that many of these shops didn't invest in terms of the capital

equipment to keep that going. Okay. So a lot of stuff that goes offshore is the really easy, simple stuff for the most part. That's always true but largely that and then maybe stuff that isn't as ip sensitive or itar more military aerospace things that just have to remain here in the united states for the most part but uh and i tell anyone if you're doing even the simple stuff you should you go with osh park you make us one of their yes yeah we do we're one of their partner shops osh park has been a great partner. James and Drew and that whole team is phenomenal. And the community and whole ecosystem they've built up is something to really, really admire.

So, $350K drills, that is a lot of money. And you said it's high margin, but you're doing other things with this money. Yeah. Yeah. So now one of the things, a few things we're doing with this, right? As far as Royal goes, we don't really have a financial partner in that regard. We've pretty much done everything off of our own balance sheet. So we don't go off and raise, we haven't gone off so far and really raised a lot of money and gone to do maybe some of the activities that you see are popular with more software startups and things like that. We've been able to grow at a pretty rapid rate and make some big bets, largely with acquisitions that we've done. The first big one being Royal itself, which you could say it's 100x the investment over the last 12 years since the company was purchased with less than 15

employees and has over 100 today and generating far more revenue per employee than was before. But we recently acquired a company about six, seven months ago in Southern California. I'm actually sitting here today. The company is called South Coast Circuits. They are a great shop that has been in business for the last 20, 30 plus years, doing a lot of really complex work for some pretty impressive semiconductor and aerospace customers. And we've been able to come in there and help grow the business and add some actual flex work and things like that from a previous company that we had acquired. And now we're kind of finishing the merger here. Advanced Assembly is another group that we've been able to grow. And then kind of the thing that I'm more involved with now is on

the venture side, something that we're new to, we're starting out with, and we don't necessarily, I wouldn't refer to us as a VC firm in that regard because we're not really investing other people's money. It's purely ours, more of a corporate venture arm. And the goals are a little bit different than that of a traditional venture capital firm in that we don't necessarily, it's more, we always think about, in addition to, yeah, the return of the business, of the startup on a purely financial basis, we're also thinking about what's the value add to our customers at Royal. So is it a piece of software that we can offer to or sell to or whatever to electrical engineers, because that's our customer at Royal Circuits. InspectAR, the company that I helped co-found

and operate as a CEO, was one of Royal's first investments. But before that, Royal had also invested in a company you may be familiar with, Snap EDA. Natasha Baker is the co-founder and CEO. She's awesome. And she's done a great job building that business over a pretty long period of time that a lot of startups don't survive this long. So she's done a phenomenal job with that company. So we invested in that company years ago, and she's done a great job building that. And then InspectAR was after. We recently invested in a company called Luminant Surgical, two kids out of Stanford. They met there. One of the co-founders, Eldrick, was actually my intern at Taser. And then he went off to do LiDAR stuff as an EE at this incredible startup. And then he did chip design at Apple.

And now he's doing this. And his co-founder, James, literally dropped out of medical school the final year to go do this. And so for us, we heard that. And it was like, I don't even need to see your deck. That's enough conviction that you guys are all in. And so augmented reality, mixed reality for surgical applications, pretty cool. They're doing some internal hardware. So basically, if you design hardware and we can be a value-add investor by basically subsidizing the cost of the boards and the parts and anywhere else we can help in that part of the supply chain, that's a good fit.

If you're designing software or hardware for electrical engineers, like if you're building an interesting lab tool or something like Inspectar, a debugging tool for EEs or hardware engineers, that's something that's in our realm. Or if you're doing something in manufacturing, we can actually be a customer and we can help you grow and kind of beat that chicken and the egg scenario. That's another place that we'd be interested in being involved. And so we're talking to startups literally every week and just seeing where we can help either introducing them to more fitting investors, investing in them directly. And we're generally very, very early and we're doing reasonably small checks in terms of usually a hundred to three, $400,000 checks. And then we're

also limited partners, meaning we invest also in a couple other funds that are somewhat focused on hardware and things like that because those people do it full-time professionally and they're really good at what they do. So just trying to be more engaged in the space and help out where we can because this is not an easy industry.

And honestly, people that choose to go into hardware or manufacturing and that area, when you could go into software and build something in that realm, like another social tool that will possibly blow up and do very well, is pretty admirable and something that we want to help foster. So you've been a CEO

and you're doing some venture funding now. If I wanted to start a small widget company, since this is not an idea I'm going to pursue, say I want to do VR for PCBs. I want to be able to walk around my PCB and look at all the things. If somebody wants to do that, let me know because I think it would be really fun. But it's all there. You have the CAD.

You have all the pieces. Maybe in order to debug, you actually walk through the blind vias to see what happens. Yeah. What do I do? How do I contact you? What do I do about equity and shares and all of that? How far along do I need to be? Honestly, the earlier the better, especially when you're starting out. Sometimes a lot of people have like,

I don't know, depending if you go, everyone gets different guidance from everybody. And some people go and before they've even got a product or customers or other investors, they've gone and kind of built up a very, I guess, custom corporate structure and have maybe something that's difficult with shares or this and that. But honestly, early on, it's all about the founder. People are going to pivot what your idea. You may love it. You may find that you have a better idea three, four months down the road.

We're investing in the founder through and through. The idea is not flushed out yet. You don't really have users or customers. It's hard to say if it's a good idea or if you're the one to take this idea to the next level. But if we have conviction in the founder that they are a smart group of people and that they want to go out on their own and actually try doing a business and are mentally prepared for that, then we'll work it out with them. And the other thing is we have a pretty good legal and accounting CPA team that

we provide to all the startups that we work with in terms of helping review their documents, come up with a favorable structure, and basically work with the founders. Because especially because we're not maybe a traditional VC firm, and by the way, many VC firms are super, super friendly and helpful and far more knowledgeable than me on this stuff. But I guess our lack of incredible knowledge on a lot of these tax and accounting structures almost helps us because we don't want to deal with the drama either.

It's like, look, we want to give you this money in the fastest and simplest way possible so you can get to work and not focus so much on dealing with the lawyers or playing lawyer yourself. So we'll work with them. Saying that you invest mostly in the founder is kind of hard because there are a lot of folks out there who are really smart but aren't good at the talking game. How do you figure out when it's a good idea, but the founder isn't ready to be a CEO? Maybe ready to be a CTO, a technical person, but isn't ready to

be a solo founder or a founder with only engineers? No, that's fair. And that's a tough one too, because I don't want to be someone who's, I don't want to be saying that you have to be extroverted or you have to be able to talk to be successful in the startup. That's not true. And many incredibly successful startups, they've been led by people that are not great talkers. I mean, in many ways, I don't think Elon Musk is the most eloquent speaker out there, but he's been able to rally people behind an incredible mission. And the value proposition for what he's doing is very succinct and it's very clear. And I don't think that requires someone to be a great speaker.

It may require a little more thinking or thought processing in terms of how you convey the message. But I think anyone that's going to be able to build a product or is technically capable is certainly capable of thinking of, okay, how do I explain the value of what I'm building succinctly? And then maybe over time you could build a sales team or hire someone that's more outspoken to go and maybe deliver the message in the early days to customers. But especially if you're selling a technical solution, I don't think you need to be extroverted in many ways that may hurt you because your customers may not be the

same way. And it can be a bit jarring or I don't even know how to say it. It's like if you're walking into Tesla for an interview and you wear a full suit, it may look a little silly because the engineers there are like, it doesn't matter how you dress or this or that. We're purely evaluating you on a metrics basis, like how you perform. And I think a lot of customers, certainly for technical applications, are the same way.

They're like, just show me the product, show me the value. And if you're somebody I can talk to about the technical aspects, that's great. So we're not so concerned with people being extroverted or great talkers, more like at least at the very least, can you explain your idea simply? And do you get excited when you explain it? Yeah, exactly. And usually you can tell. People get excited in different ways too, right? Not everybody is like me, is extroverted. And I think we've gotten a little better at

recognizing that it's some people don't get louder or jump for joy. They just maybe talk a little more. They, or they go deeper or they, or they ask questions about how we may want to interact with the product or whatever. And we're open to all those things. So honestly, it's, it's just got to be something that you want to spend the next few years on, or at least you're mentally prepared to go into business on your own for the next few years and are okay pivoting if the idea doesn't work out for whatever reason.

Okay, I've got two words for you. Yes. Electric ketchup. I don't even know how to respond to that not not good like like maybe dispenser or like when you when you're i'm the idea guy i other people need to roll with the details yeah that's fair i think that's good sometimes you need an idea guy just throw things out you put electric ketchup on your lio batteries and they suddenly stop blowing up. Oh, I see. You guys have done some interesting projects, I take it.

Nothing along those lines. You mentioned along the lines of people not being great talkers, you said you kind of assess when people aren't ready. And that can be separate from the idea is not ready, right? And how do you, what makes you think, okay, this person isn't really ready to do this yet? Well, I think if you're talking, I guess it depends on the context. Sometimes if someone's coming in saying, hey, I want to raise money. This is what I'm set up to do. I've thought through it. And then if you ask, this happens more often than you think, if you ask some, not even probing questions, simple questions about, okay, like, you know, walk me through how you may want,

how you're going to build this, how you're going to get your first maybe a hundred users. If it's a tool that, you know, maybe that's like a reasonably small number, if the tools is something like that, or walk me through, you you know what's your next hire that you may think you want to bring on or how much is it going to cost to get there this and that if those answers are tough to get to then sometimes it's like well maybe this person was more enamored with the idea of raising a lot of money but hasn't really thought through the business and maybe they don't need to raise a lot of money to to to that next stage. And actually it could be better off. They may be better off just going and building because it's not too expensive and coming back when you have something more substantive on that front. So it

just varies. Then you ask about the market, but I think what we do is we basically just try to figure out who is the customer. And then as quickly as possible, I try to go and talk to the customer, ideally in a setting where that founder and the customer are on as quickly as possible i try to go and talk to the customer ideally in a setting where that founder and the customer are on a call whether they know each other or not um we actually flew somebody in from amsterdam a couple weeks ago super super smart uh founder he's a chip designer wants to go into that space doesn't have a fully fully fledged idea out but has a bunch of interesting areas to explore and that was enough where it's like you don't meet a lot of people in this space that want to go into business on their own that have

this fire behind them that want to do it so for us it's worth bringing them like that out to visit us and let's just dive into the business. Let's connect that person for like three days of intense calls with potential customers, potential, maybe co-founders, potential employees, and flush it out. And at the end of those few days, we were able to come to a pretty interesting conclusion on where this person may want to go with the idea, when they may be ready for funding. And I don't think it was at that moment. Just having something a little more flushed out before you take on other people's money and you kind of have that added pressure on your back, no matter who it is. Which is better to have when you're seeking funding or seed funding like you do a functioning sales funnel some actually bringing money in or a committed group in the

open source or making communities i'm not sure it's mutually exclusive in that regard but i don't think the money matters early on at all because you're not going to go and at the end of the day three years down the road you're not going to survive on $10,000 annual recurring revenue, right? I mean, if that's what you got early on, that's amazing traction, sure. But it's hard to benchmark it versus if you got customers that love you, that can vouch for you, if you got like five or 10 in the first few months, that is an amazing metric, I think. Again, it depends on the startup, depends on the business, of course. But if you have users and people actually using your product and they really seem to be liking it,

you could find ways to monetize that down the road and do add-ons and whatever. And you may never even monetize, but it'll be valuable to another company as a part of their product suite. So I don't think the money matters so much early on in terms of revenue. When compared to actual, what are the users saying? And do you have users and do they love it? Or are they getting towards loving it? In a similar vein, if I brought in, if I was getting, say, 10K of engagement from whoever, is it better if that 10K came from professional companies or communities of individuals?

Oh, that, again, product dependent. It depends on who you're selling to, really. There's business-to-business products and there's companies that have, right? A lot of these kind of API software companies have purely been to developers individually or to open source groups and they've done incredibly well so it totally depends on the company i don't think one looks better or worse than the other i maybe is what you're getting at sometimes like there's that i guess that distinction or that oh this is a professional company it's more serious and open source maybe isn't. And I think, is that kind of what maybe

some people might be thinking or what you may be alluding to? Yeah, that you can't make money just from people. You have to be talking to companies or there's no real money involved. I mean, I'll give you the best example is a company we talked about earlier on the show called Oshpark. It deals with individuals and they do very well and people love them. They have really succeeded and hit product market fit in that regard. It's not like they have a new product, this and that, but the service is that good. And their customer service is awesome. The community they've built is awesome and they are tailoring totally to individuals. Okay, switching topics a little bit. What should I look for in a non-engineering co-founder? Ooh, that's a good one. I think the first one is a willingness to actually dive in

and understand and almost empathize with what the engineering counterpart is doing. Like a desire to understand and not just say, hey, it's over the wall. You build in the dark room in the dungeon. I'll go out and be in the limelight selling, right? And is that the same thing that you would want a more business side person to look for an engineering founder, an engineer willing to explain what they're doing and not just saying, go make me money? Yeah, I think that's totally so important. Like the best experience is the one that I had

that I can talk to at least. And that was with Inspectar. My co-founders, Nick, Liam, Matt, Daryl, these guys were so, not were, they still are incredibly bright, but they always, and maybe it's just because they're also very passionate people and they're very proud of what they still are, incredibly bright. But they always, and maybe it's just because they're also very passionate people

and they're very proud of what they're doing, but they always took the time to explain things to me on some of the deep technical stuff that we were doing with the augmented reality side of things, or even how we did a lot of our infrastructure and some of the safety with the users and the heavy kind of loads that these tools were taking. And it just made the experience that much more fun. And I felt like we all trusted each other when it came down to timelines and,

hey, like, when can we get this feature out? I never felt like anybody was sandbagging or trying to be like, oh, yeah, it'll take eight weeks when it really would take two. I never, ever felt that way. Everything felt super honest. We never really pushed each other in a way that felt uncomfortable. And it was just a really good, healthy relationship and everything between us. That was, in my opinion, a near perfect experience where, yeah, we're all technical in some regard. But as far as that product and that company in particular, I was not on the technical side. I was not doing the software development for the tool, but we all were able to understand and explain timelines and when things are going to get built, how they were going to get built. When we had engineering decisions, we made them as a group and I never felt excluded.

And on the same side, when we met sales and marketing decisions, we included everybody, at least on the founding team, and how we were doing things and what that might lead to on the engineering side. And it just, it made us feel like a team. Early on, especially, I think that's very, very useful. It was pretty amazing how Inspectar went from a senior project at, to a product that was being sold to a company that was purchased by a very large player in the market. Yeah. It really does seem a little too perfect.

No, really. Was it as smooth as it looked? Oh, no. And that's the thing. You don't see all the things in between like you know we'd go on an incredible high where every day we have three four five six demos and in this niche space when you have like six demos in a day where it's everybody from a large aerospace company to a massive manufacturer to a consumer phone company that is probably the most famous company in the world and these companies and this is not like a oh

this happened once in a blue moon this is every day and then you'll go two weeks and you're not hearing anything from customers and the products that are in the customer's hands are breaking you're getting a lot of those emails and you're spending all your time fixing that and being on calls you're not necessarily getting paid for that so there's a lot of those emails and you're spending all your time fixing that and being on calls. You're not necessarily getting paid for that. So there's a lot of ups and downs on the product side, on the sales side, even on the employee and the team side. And what, probably half a year in, that's when COVID hit. And so we had to go remote, but we were already kind of remote team. So we were able to kind of pivot quickly. And it was just a lot of these things that we didn't anticipate or build into the business plan that had to happen. But it was an amazing experience how honestly, just the strength of the founding team and the trust

that we had in one another to just kind of make critical decisions for the business in terms of how we're going to lay out the product, how we're going to sell the product. And then going into actual the M&A part of things, who we're going to talk to, how we're going to do it, who's going to be involved in certain conversations, and leave an ego out of it. That was a good thing that we had together, and we were able to kind of move quickly. Liam actually asked me a question for you. What do you think are some of the easy ways to make money in the hardware slash electronic space that people most often miss out on? How do I make money that nobody

else has noticed except for you? Please tell us on this public show. If I knew that one, I'd be doing it. No, it's interesting. So there's, as I think about it, some of the big trends, manufacturing, of course, at the prototype stage, but maybe a little bit beyond. There's companies like Instrumental that there's not a lot of players. manufacturing in general is an interesting space in hardware because most of these companies are still what we call job shops. They're not these incredibly professional Foxconn type. I guess I really haven't been to Foxconn, but you can understand the comparison I'm trying to make. They're not these incredibly institutionalized or professionalized organizations. They're still job shops. Everything is custom manufacturing. Every order is dealt with on an order by order individual basis. You know, there's, yeah, there's duplicity across the shop, but some things have one operator that

is far more skilled than the other. And if he's out sick, you do get a variance in the output product. There's a lot of little things in manufacturing that are messy, that are sweaty, that are dirty, that I think in a lot of ways software can help fix. And that's such a broad generalization and something maybe a very Silicon Valley thing to say. But there's a lot of companies that are doing some interesting things in software, but I think there can be more because there's a lot of potential customers, especially in manufacturing these job shop types, in electronics manufacturing, sheet metal, plastic, stamping, things like that, that are still owner do think software can come in and help in terms of everything from the ERP, but also how these machines are talking. Are they up or are they down? Basically assisting the general management and the runners and the production managers on the floor. So the manufacturing side of things can only be improved. And there's some interesting startups. Another one is I have a friend in LA, Karin, running a company called

First Resonance that is more, I guess they're also selling a lot of the OEMs, but they're doing very well in the manufacturing front. And they just raised a big round as well. And it's something that people are looking more towards. So it's something that I think if there's an interest and you want to understand the industry and take the time, a smart person or group of people can come up with innovative solutions there. The other side is on, this is something I've been thinking about on the IC side, advanced packaging and packaging in general. And this is maybe too specific for this conversation, but chip packaging is something that's critical. And you're seeing more prototyping, especially on the open source side,

XFAB and people using these multi-part wafers, kind of like OSHPARC for integrated circuits with eFABless and these shuttles. But then it comes down to packaging some of these unique chips, and that's something that isn't really widely available. The fabs sometimes do it or outsources, but there's a few packaging, yeah, folks, software and manufacturing, software for engineering. Also, chip design software, I think, can only be improved. Cadence is an incredible company.

Synopsys is incredible. But I think there is certainly room for improvement in a lot of the tool chains that people are using because they mix and match, depending, obviously, analog, mixed signal or pure digital design. So it varies. But I think you would agree, you guys would agree with me when I say that a lot of the tools that hardware engineers in particular use, both software and hardware tools, are outdated, especially in relation to our software engineering counterparts. Would you agree with that? I think if you leave off the last clause, especially compared to software engineering, I think all the tools are bad.

The software tools aren't great. Software tools are better. There are good tools in software, but there are a lot of bad tools that people still use. but yeah yeah yeah okay i think i agree with your statement yeah or or they're just maybe outdated and a lot of the reasons because you don't at least unlike the uh and this is something that i learned when we were working with some of these uh the founders doing stuff on the chip design side i was like why is the the analog side of thing especially so heavy? And it's like a lot of these guys weren't necessarily writing everything

in Python or doing a lot of the actual design like they are doing digital design today with higher-level languages. You do kind of need it more GUI-heavy. It's a lot of buttons and you're clicking through stuff and that's just kind of how it's evolved. And those tools are deep and there's a lot of heritage and they're deep in these

companies that are doing extensive chip design. And it's not something that may change overnight. So on the software side of things, you have that. But on the hardware side of things, I mean, Saley is a good company, an example of a tool that when I used it at Tesla and then I took it over to Taser, we were using in my tool. It was just a delightful experience compared to a lot of the other tools that we were using. And I really appreciated what they did as a company. And so then my next question was, well, what about the rest of the stack? What about the rest of the lab? When is that going to be updated? And that was part of the catalyst for how we built InspectR

is being one of those tools and a visual aid. But what's next for the scopes and this and that? How can we get the data all in a single place, easy to transport, so I don't need to take my phone out and snap a picture of my scope waveform? Because today, as far as I know, that's still honestly the easiest way to do it. That's the way I do it still. That's the way everyone does it. No, no, you take a USB drive that may or may not be compatible, you jam it into the scope if your scope has one,

and then you copy the file to the FAT32 file system, and then if you can find the button through the screenshot, that is. And it isn't the auto button that resets your whole thing. Yeah. And, I mean, I'm still using TerraTerm, which apparently was last compiled in 2017, but was last actually anything done to it in 1998. I mean, there are petro-cereal terminals. It's still one of the most popular

because it doesn't have all of the frou-frou stuff. It just does what it does. It doesn't meet your, it's nicer, but it meets the, this is what I i need don't complicate it with a bunch of buttons yeah that and i do have another co program on here just so you know why are you still using serial the question no but but you know when we talk about tools in hardware a lot of the stuff you know it's i i guess, certainly with software too, but you don't learn a lot of the stuff that you end up doing on the job in school. And that's why, you know, I wanted to take at least a minute to give you guys a lot of credit, your team, for writing that book, the Making Embedded Systems book, because it really, really helped me.

No, it was a team effort. Well, I give you a lot of credit, Alicia, on the book, because it actually helped me not just keep, but excel in my job and get that first product out and everything. Because I had very little experience in embedded system design, certainly at a professional level prior to coming to Taser. And it was just an awesome book that really helped me. And so that, in conjunction with using tools like Saley, made it easy to really understand what was going on very quickly

and get ramped up. And so the next thing is, yeah, so, I mean, let's see. How can I succinctly summarize my answer? Because I kind of rambled there. It's manufacturing. These are large sectors. Manufacturing, lab tools. The design side EDA varies.

I'm not so excited about it, but I think there's some interesting startups doing things that will be helpful and they seem to be having traction. So let's see. But certainly the debug side and the manufacturing side can only be improved. I think what Altium is doing with Macrofab and these teams is pretty cool.

But something to note, every factory is different. The machines are different. Even the chemicals and the processes and the travelers, which is like the recipe for any board, is not going to be identical. And so oftentimes you really, you want to build a relationship with the factory. If you're going to be doing a lot of prototyping, it'll only help, especially as your builds get more complex and you start doing stuff like flex and rigid flex. It only helps because it's, it's custom manufacturing. Every time it's different, it's not retail. Changing subjects again, you recently had an interview with me in your newsletter. Could you tell us about that? Yeah. So first of all, the interview was awesome.

And it was the second most popular link that week after I think Bob Peace's application notes on ICs, or it was the podcast on Cadence. I don't know, but it was a close second. In any case, yeah, so I started writing this newsletter, but we're on episode, we're on issue 22. So yeah, about almost five months ago, five plus months ago, I do it every week. And it's a hardware manufacturing focused newsletters, largely bringing this community together and whether it's interviews from uh interesting people in the space and then uh

or like an inside or some interesting chart where i found some data and some weird annual report or some environmental report or something like that and we put it together into an interesting chart and then also a bunch of interesting links that are not just educational in nature. It's anything, and it's not just recent news. Just anything that I think would be interesting to engineers, especially hardware engineers, it goes in this newsletter

and it goes out to now over a thousand EEs, product managers, hardware-focused, venture capitalists, and founders in the space. And it's become a really just awesome way for me personally to get involved, but also a really cool way for people to find jobs because I post interesting jobs that people send me usually every week.

Even the links, like I was telling you, 15, 20% of the links each week are sent in and I filter through from readers every week that I think are interesting to the broader audience that may be interesting. And it's also been a really interesting way to spotlight some of the companies that we've invested in and also highlight some of the companies that I'm looking at to hopefully invest or we get, you can call it deal flow from it. And there have been several companies that have either subscribed or come to as a referral because of the newsletter.

And this is all just in a few months. And so now we're kind of putting our foot on the gas a little bit. And by we, I mean me, it's, I mean, basically just me doing this and an intern and looking to build the team and do a little more my my dream would be in the next year to have a little conference uh somewhere cool maybe one of our factories and just get anybody who's a reader or a friend of the uh the newsletter to come in but it's just been an amazing experience something that i was almost hesitant or nervous about starting just like any startup. And like with any startup, the solution is to just start executing and you'll figure it out. And I've so far figured it out. And I have a couple awesome sponsors that are helping keeping this thing growing. And it's just, it's just great.

Well, I found it pretty interesting. So I it called? Oh, right. This whole time, we've never said what it's called. I don't think you've mentioned what it's called. Right. Okay. It's called the analog.io. Okay. Yep. The analog.io. That's correct. That's the URL as well. If you go to the analog.io, you will get there to the site and you can just put your email in and subscribe. Awesome. And there will be a link in the show notes, of course. Well, I'm afraid that we have to get back to work. It has been wonderful to talk to you. I believe you said something about a discount for Friends of Embedded? Yes. So Friends of Embedded, anyone that's listening to this podcast, if you just reach out to Royal Circuits and you have a board that you need to

be manufactured, just reach out to me here at RoyalCircus.com or even anyone, sales at RoyalCircus.com, and just mention that you're a friend of the podcast. We'll take care of you. There's a nice discount that we're cooking up, but you'll be taken care of. And Mihir, do you have any thoughts you'd like to leave us with? Well, this is something that didn't, I didn't believe it until it happened to me. It's often that I remind myself or have to remind myself to think about what I wanted before I got it and think about how badly I wanted it before I got it.

So the example I'll give is when we sold Inspectar, through that whole process, you build a startup, you want to get to an exit of varying degrees. And at that point, that was the best option for us and one we're very, very happy with. But you think that that's the pinnacle and you're getting there and it's in that realm of what you're working on. And all you're focused on is the exit when you're going through it. I just want to sell this company. I just want to get to the end. We're there. This is so exciting. Let's get there. After this, I don't want to sell this company. I just want to get to the end where they're like, this is so exciting. Let's get there after this. I don't need to do another startup. I'm good.

I just got to check this one off. Let's get there. Then it happens. And the year after and certainly COVID and everything else had to do with it, but it was one of the most dull years relative to running and building a startup with some of the smartest people that I've ever met. The fun is in building, the fun is in the journey. And that was a big part of why I wanted to get my pilot's license and things like this, because I didn't have another great startup idea right after that. And it almost put me in a bit

of a slump. I'm like, what am I going to do? What's next? I don't want to, I can't just go work a job in that regard anymore. I want something as stimulating as my last experience, but you got to get started to do it. And so fine. If I can't come up with something, let me push myself, do the triathlon, do the pilot's license, something to keep my mind engaged and keep going for goals. And it also puts me in a better mood. It gets my mind thinking.

It puts me back to work. And so maybe a lot of other people are dealing with this with COVID and the isolation and varying degrees for personal situations. And it's not all lost. Just keep going. Find something to stimulate yourself.

It doesn't have to be a massive startup or this or that. Embrace and cherish the wins when you get them, whatever they are, because they totally matter and they totally add up. Our guest has been Mihir Shah from Royal Circuit Solutions. Thanks, Mihir. This was fun. Thanks, guys. It was really awesome. I really want you guys to come down to the factory ASAP. Thank you to Christopher for producing and co-hosting. Thank you to our Patreon listener Slack group for their support. And thank you to Christopher for producing and co-hosting. Thank you to our Patreon listeners Slack group for their support.

And thank you to Liam and Katigan for some of these questions. Finally, thank you for listening. You can always contact us at show at embedded.fm or hit the contact link on embedded.fm. And now a quote to leave you with. You know, I don't really have a quote, but I want to tell you about this VR thing. So you put in the CAD from your EDA software, and then you can have like Tron mode where you can wander around and be the electron. Or you can have like radio mode where it rains down lightning bolts so you you can see where your antenna are on the, Christopher's laughing.

No, it's like, I mean, that's basically inspect our, right. I mean, with the AR and you can actually do overlays of anything. You can put image overlays.

You can take the readout from the, she wants to be in the board. She wants to have a headset and be wandering around the board and to, to find, you know, where the gremlins are hiding and then shoot them. Reach out to Liam and see if they can get into the next release. I want full experience,

full immersion. And then I want a first person shooter on top of it. And then can you make like, we need to add little things that, that shock you when you touch the wrong thing. And something really hot for all those times that you accidentally pick up the Saturn iron from the wrong place. That's the show. Goodbye, everyone.