#309 – An Interview with Stefan Dzisiewski-Smith

Dave Jones: This is The Amp Hour Podcast, recorded July 27th, 2016, episode 308, an interview with Stefan Krzyzewski-Smith.

Chris Gammell: Welcome to The Amp Hour. I'm Chris Gammell of Contextual Electronics.

Stefan Dzisiewski Smith: And I'm Stefan Krzyzewski-Smith of Bear Conductive and various other projects previously.

Chris Gammell: Oh yeah, I never would have gotten that right. I wouldn't have gotten even close with your name.

Stefan Dzisiewski Smith: You know, it's tough. You know, it's going to be something when you're driving to and from the nursery that my mum would drop me off at. My mum was a teacher, or was up until recently, and she would actually teach me in the car how to spell and say my surname. Because, you know, I had to kind of take ownership of it from a very early age. So yeah, it's been going all my life, and I imagine it will continue to do so.

Chris Gammell: Did you do one of those things where like a teacher would start pronouncing your name, and you'd be like, yeah, I know what you're talking about. It's me.

Stefan Dzisiewski Smith: I think it's interesting. So my mum was born on a refugee camp in Southwest. Yeah, yeah. So her mum joined the British Army as a nurse. It's actually an incredible story that maybe is not electronics related, but she joined the British Army as a nurse, having escaped from a work camp in Poland, and drove actually through some pretty terrible war-torn stuff in a stolen laundry truck, of all things, because she could speak Russian. They lived near the north border with two of her boys, and managed to end up in Egypt, where she joined the British Army as a nurse. The details are somewhat lost on me. How you get from the northern border between Poland and Russia with two boys and a laundry truck. That's a long time in a laundry truck for sure. I imagine that there's lots of details that I am not picking up on, and no doubt lots of help from other people on the way. But yeah, so she joined the Army as a nurse and was resettled. So my mother being the daughter of a refugee and being brought up initially speaking Polish, and then when they realised that they weren't going to be going back to Poland after the Second World War, and that they were going to settle in the UK, my mum is very defensive of her heritage. So when I was really young, I was very, very, you know, she taught me, and I would get very prissy if people spelt it wrong at all, and I would really defend it in kind of her way. And I think, yeah, there was a point, I think, when I hit secondary school about 11 or 12, when I realised that if I wanted to have a few friends, I probably needed to lighten up a little bit and let people come to me with it, right? Most people get to a point where they're like, it's important to spell your name right. But yeah, it's not good to go in all guns blazing initially, I don't think.

Chris Gammell: Right, right. And it's interesting now too with internet stuff, right? I mean, I'm guessing you're the only one with that name probably in the world. Maybe that's a pretty solid guess.

Stefan Dzisiewski Smith: Yeah, but certainly with the double-barreled. And it turns out that, in fact, most of the Dziuszewskys and the Dziuszewskas, so in Poland, your name takes gender. So my sister is Dziuszewskasmith, I'm Dziuszewskysmith. Oh, interesting. But yeah, all of the Dziuszewskys and Dziuszewskas that are on the internet are actually related to me or I am related to them. So yeah, it is quite interesting. And that level of connectivity, although there's still big gaps. So like I say, I don't know all that much about how Babcha, my grandmother, Babcha Zena, got to the UK. The details are with my mum really and it's all kind of secondhand. There's a lot of records that aren't linked in and available. So although the internet, I mean, it's brilliant at linking information that wants to be linked, right? But you can't grep dead trees. So there's a lot of stuff hidden in books, no doubt, that has more accurate details on this kind of thing.

Chris Gammell: Until the OCR gets a little bit better and the searching can happen through that.

Stefan Dzisiewski Smith: Yeah. And I think also the effort is there for people to put time and effort into documenting, archiving and sort of indexing these things in a sensible way. Maybe Google, maybe that's part of their big plan. I don't know.

Chris Gammell: Yeah. Well, speaking of background stories, we should mention how you and I met. So we met in Belgrade at the Resonate, right?

Stefan Dzisiewski Smith: Yeah. Sort of the back end of, you were there for Hacker Day Europe where Mike Harrison, who's a mutual friend, he had been giving a talk and you'd been there. And so had Sophie Kravitz and Mike from Hacker Day, a whole bunch of people. I think we arrived on the first day. I can't remember the Thursday, whenever day it was. And Mike was like, oh, you should come to this meal, you know, and we all got in and we're like, yeah, we got no plans on the first night. So we turned up there and it was just, everybody was there and it was suitably hectic as Resonate is every year. But it was great to sort of meet you and a bunch of other people. You all look like you'd had a pretty big, you'd already been in Belgrade for a while. You look like you're well settled. A little bit of party. Yeah.

Chris Gammell: A little bit of this, a little bit of that. Yeah. Well, yeah. And you guys brought up. You brought like an entire, like, I was really surprised at that event. So I knew Mike would be there and he mentioned there was a bunch of people from London, but like, it seemed like, like a big part of kind of the art technology design community though, like kind of just transplanted to Belgrade for that week. So what's your role like in that group? How did you get into that whole thing?

Stefan Dzisiewski Smith: So that, when I, so if we go back to, I studied electronics at Imperial College and when I got done with Imperial College, I realized that I didn't really want to go straight into industry. I'd been doing some work for a high voltage power supply design company in my free time, sort of in my breaks, my, my holidays. Realized I didn't want to get stuck into that straight away. Answered in a random email from our department administrator saying there's somebody at the Royal College of Art who wants to make a thermochromic book cover and animate it. And I was like, I remember really distinctly getting up. It was like the middle of the day. I was really hung over. I'd been out. I saw this email and just went, yeah, why not? So I got involved in this mad project. And with a, I remember it to this day, a little pic 16F87, a whole bunch of assembler code, because I don't think there were any free C compilers that I, you know, could get working or whatever at that time. Right. So I programmed this kind of scrolling message display system that used a five ohm, 10 watt flag heaters inside the book. I mean, the whole thing was a nightmare. It had a linear power regulator as well, because I was still at university.

Chris Gammell: Have you ever been electrified by a book before?

Stefan Dzisiewski Smith: It was, I mean, I had, yeah, like I said, I had a linear power regulator in there. So the linear regulator was getting hotter than anything else because I didn't even think, and I just didn't have time or the experience. You know, I was still recovering from all the knowledge going in and there hadn't been time for it to settle and turn into experience. And so I did this thing. Anyway, the guy I did the project said, you should apply to the Royal College of Art. They have this course called Industrial Design Engineering. And everybody goes, you know, you get journalists, econometrists, designers, people involved in activism, all these kinds of weird, disparate people come together at this course. And I know that they'd love to have an electronic engineer. And I thought, well, I might as well apply. And then I had to do something called a portfolio. And I was like, well, what do you mean a portfolio? They were like, you know, all the work presented beautifully. And I was like, I haven't really done any. So I've made a load of crap up, learned to use Illustrator. It looks terrible. I actually keep it. I may even have the original file. I'll see if I can share it at some point with you. I mean, it's absolutely awful. It was just loads of lies and ideas about products that I'd never made. And I think some really terrible photographs of a Nixie clock that I was busy making the power supply for. And maybe the very first PCB I ever made, which was when I was maybe 13, 14 at school. I had a PCB etching set up and we made PCBs. And so I somehow got in. And then I was like, I remember at the time I had this girlfriend from the States. And she was there when the letter came. And she picked up the letter from the front door. And she was like, oh, you know, I'm kind of so sorry because it was a small letter. And when she got her acceptance letter, it was a really big letter. That's right. I actually kind of ignored it for a couple of days. And I was like, holy crap, I've got in, but we've got a really big problem. I've got no money. And my parents, certainly, they've been very generous in funding one master's degree because I did an undergraduate master's straight through Imperial, which was four years. There was no way they were going to fund it. So I had to get funding for that. This is a very long-winded story. But I ended up getting funding and having to go to the RCA. And at that point, I was like, oh, my law. I was actually sick on my first day on the way because I was like, I am going to be bottom of this class. I can't draw. And lots of people say they can't draw. It's like I have the drawing equivalent of dyslexia. And that's not meant to be in any way offensive to people with dyslexia. But I really struggle to express my ideas visually. Very hard for me to take an idea in my head and express it. And I work really hard on drawing. I try and draw at least once a day to improve that. But I'm still terrible. Drawing is a lot about practice. And it has improved. But I have to really work to communicate, right? Because I work with lots of visual people. And if we don't have a kind of lingua franca to exchange ideas in, then we're not going to do the best stuff. So I was petrified. And I nearly quit the course. I don't know how many times people listening to this who know me will remember the day that I stormed out of the place and went to get – I just walked straight across Hyde Park, across from where the Royal College of Art is, and ended up in Edgware Road eating a little cup of kind of sweet corn with flavoring on it. Just sad, really sad on my own. And I was just really wondering how I'd ended up on Edgware Road eating sweet corn and feeling upset with the world. But I slowly came to realize that there was a lot that I could learn there. And there was a lot I could give. And there was a lot that I just had to work really hard.

Chris Gammell: Yeah. So tell me more about that experience. So you apply to this program, but is the expectation that you basically go through artistic training like – like this would be like the equivalent of like an MFA-type program in the States, like Master's of Fine Arts type of thing? Or what is it?

Speaker ?: Sort of.

Stefan Dzisiewski Smith: So, I mean, it's at the Royal College of Art, which is a postgraduate-only institution. But the degree when I took it was – so they are all masters. The degree when I took it was a master's in industrial design engineering. I think it's now referred to as innovation design engineering. But the idea was that you could come from any background. As I mentioned, there are lots of different people from disparate backgrounds. And that you would come together and work on design-oriented projects. There is some technical guidance there. So there were drawing classes and there were classes on how to do 3D modelling and additive and subtractive rapid prototyping processes, design thinking. There was a lot of peer-to-peer learning within the group. So if you – there was a lot of group work as well. So let's say the very first project we did, we had to create actually the only artistic, purely artistic thing that we did while we were there, which was an art intervention within London. And so we ended up making a whole load of origami birds out of cartridge paper and sticking them up in a tunnel that goes underneath the railway that's used by buses in Vauxhall Station near Oval Cricket Ground in London to brighten up the tunnel. So we had this idea that London's got lots of lovely, beautiful places, but it's also got some dark places. And we wanted to make one of them up. But the whole process of getting to that concept and justifying it – I mean, going through critiques. In the first week, we had to stand up and say what our idea was. And our original idea was something quite different. And I remember to this day, it was like the end of the first Friday. I was exhausted emotionally, exhausted kind of socially. And they told us that our idea was completely shit and to go away. And everybody who'd been through design sort of school was like, oh, well, you know, it's the first project and it's early on. Of course, you're going to have negative feedback. And I was just destroyed. I was like, it's a really nice idea. What's their fucking problem? Like, why do they have to be so mean? And that was the start of me taking stuff way too personally for quite a while on the course. But you build shots. And I think actually having the ability to take criticism, which I think people who know me will still say I'm not the best person at taking criticism. And I do take things to heart. But I at least now know not to shout back, to listen about it, get on my bike, go and ride all the pain out, and then start to deal with where the truth is in what they're saying. And whether I think I can improve or, you know, iterate on something. But anyway, so.

Chris Gammell: No, no, that's interesting too. Because, I mean, surely you had to do the same kind of thing for like design reviews for schematics and the technical stuff in the past. But somehow it feels like there's more, that's more of a logical thing. It's like, oh, I was wrong here versus, oh, well, this is the thing, this idea I came up with. And now people say that's a bad idea, right? It's more, it feels more personal or sounds more personal at least to me.

Stefan Dzisiewski Smith: I think there's an element of that. I think there's also an element of within engineering, the brief tends to be tighter earlier. And it tends to be a more linear path. I've realized that there is always problem solving. There are always things to iterate and sometimes you have to rip up and you learn completely new ways of doing things. But within the engineering environment, one, I was, when I was working in this high voltage place, I was working with some brilliant, really talented and very friendly engineers who wanted, who took me under their wing. I was developing test equipment for their, for their designs largely. And they knew what they wanted. And in a way they wanted me to be kind of an extension of their arm so that they could do more work, but also to use that to inspire me to think about NICs and to take that on board and then synthesize some of my own ideas from it. So it was quite friendly. There was a lot of scope for failure and there was a lot of scope for small innovation. But the difference with...

Chris Gammell: There was room for failure with high voltage? Because that's not my experience.

Stefan Dzisiewski Smith: Well, there was room for failure in the areas I was working with. We did have a few mishaps. We did once have the floor set on fire. I had some people call up to my office and say, the floor's on fire because arts had just found their way down. We did all our testing at large cages. I mean, there are a lot of... When I said there's a lot of acceptance of failure, there are also a lot of safeguards. So I said, okay, the floor's on fire. Have you hit the emergency stop? Not yet. Why are you on the phone to me? Go to the emergency stop. Put the phone down. Stop buggering around. Big red button. Yeah. Yeah, exactly. And that was... I mean, that was an interesting one. That was where you start to find out about tracking. You know, really about tracking. We had something like an 180 kV, 125, 128 milliamp output. You know, power supply for an x-ray tube. And we were testing it on a resistive load, which is sort of a five, six foot high perspex triangular prism with resistance arranged going down it. And what had happened is they needed to do an offload test, but they'd left the... So the way you normally do that is you leave the connection at the top of the resistive load, and you disconnect the ground return at the bottom, right? So you set up 180 kV of field across the five feet. The problem was that the previous test that they'd done was a half load test where you look at the current limit and protection circuit. So you run it into half the expected load and ramp up the current until it clamps, right? And that had passed. But then they pulled out the ground return at the bottom. Now you only had half the distance of the stack and then set the output to full kVs. And the current just went, oh, let's see what we can do here and sort of snaked its way around. And yeah, found the floor and made a sizable burn mark, which I have to say was my fault for not being very explicit about saying and kind of the order of operations in the test. It was a semi-automated test as well. So this was actually running. We were actually controlling this through a PC. So we did add some safeguards subsequently to that. That's good.

Chris Gammell: So, well, I want to get back to the design thing real quick in a second, but this high voltage play. So what was the, it was like a test lab or what was it? So it was a company.

Stefan Dzisiewski Smith: I'm not going to mention the name just in case stories that I say are considered less than positive. I think it was a brilliant place to work. But obviously I've just told you there about a mistake that I made. I think we dealt with it really well, but I don't know how they would feel about me telling the story. I hope very positive. They're a lovely bunch of people. But they manufacture and develop high voltage power supplies for a number of different applications, including a number of analytical X-ray machines, some medical X-ray and also things like power feed for both ship-borne and land-borne power feed for fibre optic telecommunications. So it was an amazing place to get as your first place to work. I mean, everybody else is working on stuff at normal voltages and I'm trying to remember that 10 kV per inch is a good tracking distance in air. And you can do 10 kV per millimetre inside rubber, but then you've got to think about the boundaries between dissimilar materials. And, you know, we had a plasma etching machine in the potting environment to plasma etch the surfaces to prime them for good keying to the rubber. All sorts of things like this that I would never have experienced in any other way, you know.

Chris Gammell: Oh, so like the plasma pitted out the surface so that when the next layer bonds to it, it has more surface area kind of thing? Or how does that work?

Stefan Dzisiewski Smith: Exactly, yeah. So you deliberately erode the top area. It's partially a cleaning thing. And then also, yeah, I think as I understand it, there's sort of a micro-pitting that then allows the rubber to key really well because you don't want inside a potted rubber stack on a high-box power supply. It's for there to be dark discharge or tracking within that starts to break down the rubber because then the rubber starts to carbonise and then it starts to conduct better. And then eventually you end up with a slow and eventually catastrophic breakdown that it's also then really difficult to repair without digging out the rubber. I mean, I definitely spent some of my time digging out rubber. That was a boring job. And the rubber's like this really dull old gray as well. So there's no fun to it. Although we used to make, you know, we always had spare rubber over and we'd make like rubber balls or rubber hands or various jigs for bits of bits. There was always kind of a competition to see who could pour rubber into what interesting mold. That was towards the extent of the fun you could have though. I mean, you did have to play sensible.

Chris Gammell: Right. That makes sense. No, that's cool. And is it layered because of like air? You need to get air through it and that kind of thing because you couldn't just pour it all the way through?

Stefan Dzisiewski Smith: No. Well, you do. You pour it up to it. So you normally pour it over the high voltage end of the stack. If you've got a board with low voltage at one end, so you'll have the forward part of the control loop and then you'll have the stack and then you'll normally have a high voltage, just a high impedance, well-matched resistive network divider. And then that will go and feed back to the low voltage end and into the feedback loop. I mean, obviously the topology depends on which parts need to be isolated from where and whether it's positive or negative. If you've got a grounded or floating filament for an X-ray in there as well. But you tend to pot just the high voltage side and you leave the low voltage side un-potted because you don't want to pot anything you don't need to. And obviously you want to be able to get at things like trim resistors and things like that. Oh, yeah. That makes sense.

Chris Gammell: Well, that is really... Yeah, that is a new topic for the amp hour, I think. I think we don't really talk about high voltage very much because, well, personally it scares the bejesus out of me. So, yeah.

Stefan Dzisiewski Smith: I got a few healthy shocks. I used to budget for one a year. I got some big ones. Yeah, I definitely got one where somebody had managed to disconnect the protective earth on the casing of a unit that basically floated up. I mean, it was another 180 kV, sort of a four kilowatt unit, which is quite frightening. And as I touched it to switch it off, because we could hear ticking, I should have hit the emergency stop, but it was late and I wasn't thinking. And as I touched the case, I'm pretty sure it was through, I guess, the X caps in the input mains filter. The return current had found its way back through there. So I got sort of probably about 5 kV. It would have limited. But yeah, about 5 kV through some caps. Very buzzy. I remember it now. I can visualize it. It makes very clear memories. But there's lots of other fun things. You know, looking for arc discharge in high voltage setups, either listening for it with really long acrylic tubes, because obviously you don't want to get anywhere near the unit. So using long acrylic tubes to try and listen for where the ticking's coming, or putting your lamp coat over your head and looking in a very darkened room to try and observe UV where ozonation was taking place. That's another good technique. And obviously thermal cameras as well. So you can do quite a lot of circuit debugging using thermal cameras without having to touch anything.

Chris Gammell: Yeah.

Stefan Dzisiewski Smith: And these are all good techniques that actually I've taken on in my career and used elsewhere for different reasons. So yeah, I owe an awful lot to everybody there, really.

Chris Gammell: Yeah. No, that's it. I mean, that is usually the... I mean, and the important thing is that, you know, you were okay. And I think the interesting thing is like what you said about the... It was late at night too. I think probably one of the biggest safety protocols is like, you know what? Everybody goes home after eight hours just because you get tired, you start making mistakes.

Stefan Dzisiewski Smith: It wasn't late at night. Sorry, I should clarify that. It was late for me, by which I mean five at the end of the day. And I'd probably been partying the night before. So yeah, I brought some of that on myself. But the safety precautions that were taken there were fantastic. I mean, everything got written up, all incidents. Everything was taken exceptionally seriously. Because we were working... I mean, we had x-ray tubes as well in lead-lined boxes that we were building. And they weren't always 100% first time. I still remember one being, you know, equivalent, I guess, of bringing up a board, but bringing up an x-ray tube in a kind of these kind of coffins. I mean, they looked like, without wanting to sound horribly morbid, a kind of coffin for a child lined with roofing lead that was then folded up into this wooden coffin, sealed at the corners with a big heavy-duty soldering iron. And then a lid on it. And then you've got a tube, which is, I guess, the ones we were using. Again, about the size of a child. I don't know why that's my frame of reference, but they really were. And I guess they were liquored. So we treated them a bit like... Well, we treated them delicately. But then you've got normally a water cooling system in there and possibly also a vacuum system to maintain the vacuum inside the tube. Some of them were vacuum caps. So you would vacuum down the tube and then cap it off. Some of them were on a continuous vacuum pump. And then you've got water cooling to stop the, well, stop filament melting and stop the beryllium tube exploding. Sorry, the beryllium window on the end of the tube sort of exploding

Chris Gammell: or doing anything like that. And I remember... So the lead line box, that was just to make sure that if there were any emissions, it just... Oh, no.

Stefan Dzisiewski Smith: There absolutely would be emissions. We ran them live. We dumped significant load into that box. The thing is that X-ray tubes are very difficult to simulate. So we did have some tube simulators that simulated by taking the low-voltage control end of the circuitry and feeding them through and trying to emulate. Because what you've got is you've got a filament. You're running the filament in a closed control loop saying that allows you to control the current through the filament. Then you set up the voltage using your voltage feedback network through the resistive divider. But then you set the emission current from the tube by adjusting the filament current. The voltage and the current are independent from one another, relatively speaking. But you're...

Chris Gammell: So it's like a valve, right? Like a vacuum... Like any other vacuum tube, right? Yeah.

Stefan Dzisiewski Smith: It's just a hot filament valve, but the target is... Yeah. Yeah, it's something different. Normally tungsten. Yeah. And then there's other bits and pieces, things like micro focus in there as well, electrostatic deflection and all sorts of crazy bits and pieces. I mean, the people who worked there were phenomenal, absolutely phenomenal at what they did. And I don't claim to understand even one-tenth of what they were doing, but I got to see a lot of it and to watch some really professional people make what could have been a very dangerous environment exceptionally safe. But yeah, we had a day that one of those tubes we were going... It was ready... The box was ready to be checked and the chap who was doing the sign-off was going round so you set high KV and low milliamps initially, so you've got very good penetration, very few particles going through X-ray photons and just going round with a scintillation count around all the edges and one of the edges hadn't been soldered. It'd just been overlooked and the scintillation counter went crazy. The emergency stop was hit very quickly and we all looked very worried. But the procedure was in place to be very sensitive at a very low level. We had dosimetry badges both in the area and on every person

Chris Gammell: and we were all

Stefan Dzisiewski Smith: quite interested in what might come back at the end of the month but it was barely a blip. We experienced far less radiation working there than you probably would anywhere else just because we were super tight on everything.

Chris Gammell: Right. You get more in an airplane, right?

Stefan Dzisiewski Smith: Yeah, you do get significant amounts in an airplane. Obviously, there's stories like don't take Brazil nuts or bananas into a nuclear power plant because it'll set things off.

Chris Gammell: Yeah.

Stefan Dzisiewski Smith: But yeah. And the one that I think really gets people is people who work on scrap metal sites finding interesting pieces of metal and picking them up and putting them in their pocket not realising that they're obviously not an x-ray generator but they're a radioactive substance and they've just placed it very close to their skin. There were some really horrible pictures that our x-ray when we did x-ray training days, radiation training days that they showed us and that can go real bad real quick.

Speaker ?: Yep.

Chris Gammell: Yep. Yeah, I've seen some of that stuff because well, not former guests but Jeff Kaiser who's on here all the time he has that that Geiger counter tube project thing. Yeah,

Stefan Dzisiewski Smith: that's the open counter, right? Is that the one?

Chris Gammell: It's not the open counter but a lot of them are the same. They all use those old Russian Geiger tubes basically so they're interesting interesting projects.

Stefan Dzisiewski Smith: Yeah, they're super interesting. I mean, I am so, I mean, this kind of links back to Mike Harrison. I guess I should let's well, it's up to you but we could talk about how I know Mike because it kind of ends up it ends up with me buying Mike an airport x-ray scanner for his YouTube channel. Aha, so you're the one, huh? Yeah, that was so Mike is a fantastic chap. I sort of admired him from afar through his website when I was interested in building mixy tubes and then so I finished the RCA and I wanted to get out of London. I was completely fed up with all of the pressure. I kind of just didn't know what the heck my life was about anymore. This is very common amongst RCA graduates. It's a very intense pressure cooker of a course and so I went back to Brighton which is where I'm originally from. I got a house just outside or got a flat you know rented a flat just outside of this train station with some people I just met on Gumtree so some random people I'd never met before and I was working at the high voltage plate and I was I started a darts team and I just I mean I love darts I've got a few loves in my life cycling darts and cricket are my main interests. Oh my god. So British. I know, I know I should have a portly belly but the cycling helps keep that at bay but I yeah that's good yeah dude I was just working it was 8.15 till 5 every day and we only did a half day on a Friday and I was in Brighton which is a great party town and I was playing darts and not really caring about the future just living a very happy day to day life when one of the people that I'd studied with at the Royal College of Art who had previously she had studied in Brighton and then she was from London and had studied in Brighton then come back to the Royal College of Art to do her masters and I'd kind of done the opposite where I'd grown up in Brighton and studied in London and she said Stefan I used to work or I've done some work at this architectural lighting design and art installation place they're looking for an electronic engineer I really think you should apply and I went kind of yeah yeah I don't know I don't care I'm hungover I'm playing darts tomorrow I'm doing whatever and

Chris Gammell: I've got priorities here come on yeah exactly

Stefan Dzisiewski Smith: so I was very dismissive and then I got a phone call again saying I really think you should apply you know I think it'd be great this is a chance to do some really creative stuff you know work on everything you've done at the Royal College of Art and I wasn't really sure I wanted to I wasn't sure of anything and then I got a call from the managing director of the company saying we'd really like you to apply and I was like yeah I'll get around to it and then I got a call saying you're coming to just get to London come now do it so I went to see my boss did you have the job? well yeah I did I went to see my boss at the High Voltage base and I said look Matthew I'm going for an interview and I think if I get it I'm going to take the job and he was like well we never really thought you'd stay here as long as you did we're happy with the work you do but you know you're young you should go and explore some stuff and I was really I was really buoyed by the kind of support from him and I think that gave me the confidence to go and see what was what and yeah I got the job and I mean I think there was a dearth of applicants to be absolutely honest which is something that continues in London there's a there are more jobs for creative electronic engineers than there are creative electronic engineers to do them which is both a brilliant and a frightening position to be in in terms of if you like sleep

Chris Gammell: it can be difficult sometimes right right right it's all feast no famine kind of thing yes

Stefan Dzisiewski Smith: I mean certainly that's my experience of it maybe I'm particularly fortunate but that's the way it seems but yeah I went up and I started working there but I went from a company with 20-30 technical engineers and test equipment designers to just me so we're going to do we're going to take half a million pounds off this city or this particular architectural installation where they're building let's say a huge arena like the Matthew Knight Arena in Eugene, Oregon and they want an artwork and what we want you to do is to work out we'll come up with some concepts that basically are physically impossible we'll take up a lot of time designing them we'll spend a lot of money in fees and then we'll give you about five pounds in two weeks to make an artwork that we've charged half a million dollars or whatever for I mean that's a gross exaggeration but I did I suddenly had to do software I'd never really written any vision stuff any computer graphic stuff I had to I learned open frameworks I bought numerous books on C, C++ just cracked onto sorting that out I mean it wasn't that long previous that I'd been writing PIC assembler code and cursing the PIC structure for having banked memory and now here I was trying to debug network cameras over fibre optic links in a stadium with doing optical flow analysis and then driving custom LCD panels with Mike Harrison's drivers so this is where I sort of really got to know Mike in that they said okay I can't remember which project I think it was actually the Oregon project was the first time that I was like look we've got to make these driver boards and I'm not really sure how to go about it I think I might need to you know find an external electronics consultant and they said don't worry about it give this guy a call or an email and it was like Mike at whitewing.co.uk I was like whoa you work with Mike Harrison so suddenly there was a kind of fanboy moment and I was like oh man so I emailed him my incredibly terrible spec for driving LCDs where I'd drawn half the circuit using XOR gates and God knows what else and Mike was just like yeah we're going to use DMX we're going to have this many per thing yeah what's the capacitance and the size of the panel and I was like I don't know and somehow Mike's infinite patience and kindness helped me to get into discussions with the LCD company I mean I'd never worked with LCDs before I'd never worked with DMX before I'd never worked with differential well I think we've done a little bit of RS485 in some of the test automation at the high voltage place but nothing on that kind of scale

Chris Gammell: well to hear Mike talk about it he yeah he just gets excited about the problems too so like that's what he chases no matter what exactly and I think I'm sure that you not understand you're coming to it with a real problem is probably the best the best case scenario if you were to come in with here's how we're going to do it that's a different thing I think

Stefan Dzisiewski Smith: yeah and I think Mike appreciates people who listen and learn if you're not going to pay attention then I think he has less time for you I think he he likes people who put the effort in on their side as well which is reasonable and also I think we were very lucky that the LCD shutter panels that we were using weren't something that he'd driven before and you know he had a lot of the prior experience to do say the DMX back end but the driver that he was going to onto it was something different and yeah and that's how I got to know him and I worked at the this artwork place for I guess two and a half years before really just burning myself out pretty terribly but I learned an incredible amount and I made some amazing friends excuse me the majority of who were the people who turned up at that dinner so there were probably I think we started as about 12 I think by the end of the Belgrade tour by picking up people who you know friends of friends and people that we'd worked with there were a group of nearly 20 of us marauding around Belgrade but in a very genteel and loving and respectful manner

Chris Gammell: of course of course yes right right right not shouting well no

Stefan Dzisiewski Smith: I have to be very clear British people don't have the best reputation for holidays abroad in terms of raudiness so we must be from some of the people in Belgrade yeah exactly well look I think you know it's very difficult especially with the UN intervention in the in the Yugoslavian war it's that's very recent and very real and it's not I think there are certainly taxi drivers there who were very open with me that they weren't impressed of the idea that British people just come there because it's cheap and when I explain right right I think yeah you know I explain that we love the city and we we come back and I mean we definitely unlike the last night after Holly Herndon we ended up in this bar just having a few you know liveness before we we started to chill the night out and there was a chap who got really kind of shirty with my friend Martin and started talking about the fact that I think his the gist of it was that in some way Martin was responsible for the bombing of the city

Chris Gammell: oh my and it was a really

Stefan Dzisiewski Smith: tense moment actually I I was quite impressed the guy's friends encouraged him that they should leave and leave us to it which was a very pacifist way of dealing with the problem because I don't know what we'd have done if it had got any more difficult and there was quite a big group of us we would have been difficult to motivate and explain to each person in the group why we now had to leave although Martin had done nothing wrong so yeah there's tension and I think that's something that you learn to deal with when you're working or visiting other countries there's always going to be prejudice and bad reputation and things like that and you just have to work on being a decent person right and friendly and open

Chris Gammell: yeah gotta gotta respect the local culture too like we were talking about before the show exactly when I go to when I go to London it's not I don't call them chips I don't call them potato chips I call them crisps exactly I don't call them french fries I call them I call them chips you know you gotta respect the local tradition yeah

Stefan Dzisiewski Smith: and I when I was working in Oregon I mean I had a great time working in the states we did a number of projects in the states and the Oregon project was actually fabricated in South Boston so I did quite a few really yeah which is a super cool really really nice signage company there who just took on the fabrication and were quite happy with the fact that it was really weird you know from their point of view the kind of most interesting or sorry the most unusual stuff they might do was a large three-dimensional steel and aluminium baseball for a baseball ground and suddenly they were doing a structure it was all epoxied together I'll have to find a link for you to share but this thing was absolutely huge I seem to remember something like 12 metres by 7 metres so 20 was that 21 feet by 36 feet and it was all like epoxied together I'd never seen so much epoxy going into something buckets and buckets of epoxy but they knew what they were doing you know exceptionally professional I will say that my experience of American fabricators and American workers in general is that they have much more of a can-do attitude than some British I guess the average level of can-do attitude in the states to me seems to be higher in industry than my experience in the UK I'm very lucky to have worked with some really brilliant people but I do still work with some suppliers and subcontractors who you almost feel like you're forcing your money into their hands and encouraging right whereas

Chris Gammell: I don't think that's a regional thing I think it's more of a when there's competition anywhere you're going to get people that you can go seek out the best ones and then in the situation you're talking about here I would imagine if you approach a signage company who does good work and is normally doing day in day out reliable work but then you approach them with something weird and interesting like that they're just going to double down and do awesome work on that I guess

Stefan Dzisiewski Smith: there is an element that if you are open to inquiries from people from another country who you've never heard of before asking you weird stuff then you're probably self selecting as slightly more open minded and talented people so they were absolutely brilliant to work with I had a great time doing it but yeah it ended up being too much for me I really was working something like six and a half days a week and that's not good for anybody I mean it was an unhealthy lifestyle

Chris Gammell: like the high voltage thing well yeah

Stefan Dzisiewski Smith: and potentially as well I mean I had to learn how to work at height I remember dropping again one of Mike Harrison's soldering irons I was up inside a 12 meter structure with a further sort of five meters drop underneath so 17 meters what's that 51 feet and I dropped Mike's gas powered soldering iron which we shouldn't have had to be had they followed the test you know this was assembled in the UK another LCD structure but it was done in such a hurry that they didn't follow my test procedure which I said please test all of the modules before they come to sight well it wasn't completed and they assembled it in a rush and trapped wires so I was up there soldering dropped this gas soldering iron and it just went straight down 51 feet and bam on the floor I mean it didn't explode but I had to buy Mike a new iron but that's definitely a result of I'll be honest I had not had any proper rigging training I was in there with another rigger right they were just like put on the harness I mean it was a seated

Chris Gammell: harness we'll tie you to something yeah it was kind of like just go

Stefan Dzisiewski Smith: for it and I was shaking I was really shaking Michael tell you he was on site Mike was up and down in there like a you know like a rabbit up a drain or rat up a drain pipe rather but I was yeah I struggled it was it was quite funny but we got the job done I don't I don't I don't do heights I don't do heights personally and it was swaying around in the wind now I'm not doing that again nope no it was claustrophobic it was only like sort of four feet wide inside oh yeah another bad memory

Chris Gammell: so so I'm interested I'm always interested so I you know I've talked to Mike about a lot of his projects and I I he came and gave a talk at one of my events out in San Francisco as well and he was talking about some cool projects but the thing that always interests me so like you mentioned the LCD manufacturer so like those things don't come with drivers already on them the manufacturers are just basically giving you raw elements I always assume that like for all this stuff it's like oh well there'd be drivers attached you know what I mean like because you can buy small versions with stuff already integrated but that's not really the case for your stuff

Stefan Dzisiewski Smith: well that yeah I mean this particular example I guess the most similar thing the object that it's most similar to was a an automatic welding mask and because it's just a single cell it's really not that tough to drive you provide you know a symmetrical AC waveform of a few volts say five volts peak to peak or sorry no five volts amplitude and that's enough to drive it and make it go opaque and if you remove the voltage drive it back to zero it drains the charge out of the inherent capacitance and it goes back to being clear or vice versa depending on the polarization

Chris Gammell: oh so this is this is shutter glass stuff like this is the one that was in there yeah so

Stefan Dzisiewski Smith: it was sort of two and a half thousand shutters but with ink inside each of the panels so that was actually right we brokered that through a company in Ohio I think LXD LXD so I think LXD might have moved out of Ohio or our contact there might have done but yeah the toss up was either Kent displays or LXD and I know that they farmed the manufacturer to China but they were great in developing stuff and I certainly learned so we had to color these shutters in green and yellow which were the colors of the Oregon ducks go ducks yeah and I initially just peeled off the the polarizer off the back of this LCD the sample that we've been given and put a lighting gel in there and that wasn't good enough it just kind of broke it because the lighting gel was depolarizing the light and so you you have to have a non-depolarizing uh pigment in there and I sort of said oh well can we get hold of some and 3M do it and they do everything but you have to buy like an insane amount um right like rolls and rolls and rolls of it kind of being uncut it actually comes as a liquid I think and you pour it in um during oh really yeah I don't know the details of the manufacturer that's what um LXD did for us but they yeah they were brilliant with it um and it was really interesting to sort of see and get involved with but yeah I I did that for for quite a while I worked on some stuff for the Olympics um I got to go to some great parties uh I got to do some some cool events and some I was on some TV sets sort of hiding out of the way uh in in the sort of shadows whilst things whilst television adverts and things were being shot and it was so you're like the man

Chris Gammell: behind the man behind the curtain kind of thing yeah about four layers down the man the man crying right right right right the soldering iron and the salty tears exactly uh I still remember

Stefan Dzisiewski Smith: a particular film director who was shooting an ident for channel four one of the main terrestrial channels in the United Kingdom um who had got himself a laser pointer to point out parts that were not functioning correctly in this thing that we'd thrown together in an inordinately amount a short amount of time and started filming at 5 a.m. and I was exhausted and he was just going here this one this one and uh I was fixing one he goes hurry up and I turn around and we go I'm sorry fella I've got two speeds I've got how fast I was going and really fucking slow so it's up to you it's up to you you can't make me go faster because it will not work so you're gonna have to calm down I understand this is your thing but you can't just start getting lippy and um he went off and got a coffee uh and I got some space to work but I yeah I don't think I dealt with that as best as I could but that's what 5 a.m. me was feeling um I was not impressed right uh so yeah I'm not always great at client relations uh I'm getting better yeah I'm getting better but yeah

Chris Gammell: so okay so you so you're not doing this anymore like surprising that that sounds stressful uh but what so what what do you have you been doing in the

Stefan Dzisiewski Smith: meantime so I finished up there and I well I I kind of put my foot on one boat while I was still there so I I got a call from a friend again from the Royal College of Art from the year uh ahead of me and he said look um I've got this project with a bicycle generator to charge mobile telephones in Uganda uh do you want to come on board and I thought oh I don't know about this magnetic power generation low voltage not even high voltage mobile phones hot countries I don't know if this sounds like a good idea but again because I uh I'm one of those people who likes to to take on silly challenges I went all right and well come on then let's give it a go he goes all right we've got a bit of funding we're going to be uh in the innovation RCA incubator um I need you you know to come and take a look and then we'll sign some contracts and we're going to try and make bicycle generators to charge phones so I arrived uh on my first day there having said goodbye to everybody at Jackson Brew studio um although more kind of au revoir because we still hang out a lot but um I arrived there and they had this bicycle uh on a kind of frame and a motor strapped to it a DC permanent magnet motor and they were saying we just want to ride this charge a battery and we'll sell it and people will ride the bike from town to town and they'll charge their they'll the person who's riding the bike will stop and offer people charges and put it up on this stand and then it'll spin the wheel to generate power and I said okay well we'll see what we can do uh in terms of power and a number of phones that we could charge and how much we'll be able to sell it for and it was pretty obvious fairly quickly that it was going to be a tough ask for one thing for maximum power transfer you're looking at 50 right to get if you do impedance matching you're talking about 50 efficiency transfer uh electrically so you were potentially throwing away a lot if you wanted to do um impedance matching uh-huh you were also talking about quite a lot of peddling to generate enough power to charge up a battery to charge enough phones to really make a business um the charge to uh it costs almost no matter where you go in the world if you're having to pay charge by charge whether you're in india or uganda or uh kenya and you're having to pay to charge your mobile telephone uh it costs about 10 pence roughly so i guess i don't know what the conversion to dollars would be since we decided that we're leaving the european union oh yeah you're our first uk guest since that's happened that's like a one cent now right i don't think

Chris Gammell: no 10 pence would be uh i think like 70 cents so no sorry

Stefan Dzisiewski Smith: seven cents seven cents right something like something something in that range anyway i've forgotten where we are with that uh but you need to charge quite a number of phones in order to make a day's wage so a day's wage there oh you can live reasonably for about two pounds so you need to charge at least 20 phones with no costs well actually that's a lot of peddling in a very hot

Chris Gammell: equatorial climate right um right i think yeah people always lose that too they're like oh well it's just a phone like it seems like but like no like if you look at the power that you need to actually fill that thing like the capacities are pretty significant right i mean like it's on it's it's in watt hours right i mean it's like that's what you need to fill it up with is energy you can yeah yeah

Stefan Dzisiewski Smith: it's exactly that you you you do the energy this is where the concept of energy is fantastic right the idea that you can equate mechanical stuff with electrical stuff and then put an efficiency factor in between the transfer ratio yeah exactly i mean and just to do some modeling with that it's very powerful so we were

Chris Gammell: looking at that it's really good actually this is this is i'm sure dave if he were listening he's not listening because he's he never listens but uh he's like he's laid up right now too with his knee yeah but but like it's like the the solar roadways type stuff right i mean it's just like it doesn't doesn't quite compute unless you have a super high transfer ratio and most of the time you

Stefan Dzisiewski Smith: won't exactly and that's what we're saying you know if you're trying to do peak power transfer then you're talking about 50 efficiency if you're trying you can do lower power transfer but then you're uh you know then you're talking about longer longer transfer time so it didn't work out but what we did work out was that solar panels are not particularly costly to to purchase but there are a lot of failed solar panel projects out there because they are either a race to the bottom or the charge or the cost of actually buy them with the very high interest rates that you get um with micro finance in places like uganda and india uh is prohibitive so the finance um the interest tends to be very high because a lot of people commit to um the idea of having something and they commit to what the payment should be but actually they default on the payments and then that results in equipment having to be recovered which is a costly process that gets passed on in the interest gotcha and so what we got to was well what if you had a solar panel and we give them to people to run on our behalf and we do a profit share with them so we provide the equipment to trusted people this is a model that works for uh finance um so there are companies who will provide local banking services giving hundreds of dollars in some case to low into local uh banking operators and a kind of card that looks a little bit like the chip and pin device that you get in a restaurant when you pay the bill yep and they will manage accounts for people and allow them to pay into an account and pay out from it without them having to travel to a central more urban location there's a lot of the people who i think i'm skating over a lot of the context here but a lot of the people who have mobile telephones let's take uganda as an example will be um farmers growing crops like coffee or perhaps vanilla or cocoa or something like that and selling that seasonally so they spend a lot of the year growing the coffee in their um uh in their plot of land um or growing bananas or motoke or anything else and then they especially with coffee which i think you get normally two harvests a year you'll get a big lump of money in one go yep and you'll be super rich but it's very hard to then keep hold of that money and manage it i mean if i paid you twice a year right right your job it is harder to manage money that way of course right you have

Chris Gammell: to do projections and it's probably formerly all cash and stuff too right

Stefan Dzisiewski Smith: exactly and also you're talking about doing projections for people who may not even have high school education yeah so the understanding of just taking this is the amount of money if i divide that by the number of days in half a year that's how much i can spend a day yeah that's a difficult thing to teach people and there are whole uh charities and ngos who teach these kinds of skills to people that yeah um but a lot of these people will have mobile telephones because it allows them to do very simple market trading if you think about the shots of trading floors that you get where you've got people on telephone shouting buy buy sell sell what's the price of oil here and there yeah blah blah blah commodity stuff yep well it's it's the same thing if you've got bananas or you've got coffee to sell today you know that you can sell it in two places you can go uh let's say you can go down the road to ginger into town or you can go up the road into umbale which are two towns on a road that runs from sort of entebbe uh in uganda out to kenya and if you're between the two you pull up both of them and one says well i'll pay you this much and the other one says i'll pay you a little more then you know which way you're going and it's a very valuable way it's also a way of staying in touch it's a way of keeping secure it's a way of maintaining health amongst families so if you know somebody is sick then you can go and respond yeah

Chris Gammell: isn't it also aren't they starting to do like micropayments through phone mobiles

Stefan Dzisiewski Smith: as well there's a lot of that yeah so uh schemes like mpesa um and uh mtn mobile money uh there are different schemes depending on the the mobile telephone network the mobile network provider uh in that country right um so yeah it's a way of sending money between people and the banking as well so there's huge huge benefit um to having a mobile telephone if you think that you're in a place where there is no functional postal service outside of the post box in major sorry the post office in major towns yeah it's very difficult to transport things everything goes dark at night pretty much um and it's a long distance if i said to you okay you can charge your mobile phone but you've got to travel for an hour and a half so what's what's an hour and a half from from uh cleveland uh erie pennsylvania so you can only charge your phone in erie pennsylvania right let's play that game right that starts to get quite annoying and i bet you wouldn't charge

Chris Gammell: your phone every day and i also bet that i wouldn't be able to drive my car there would probably be like taking a bus or something uh some slightly like yeah you'd

Stefan Dzisiewski Smith: likely be either getting a lift on the back of a motorcycle yeah um or getting yeah the kind of local taxi buses which yeah right until they're full and then then go between places and it costs you money to do so right so actually convenience is really important as well convenience is a a you can put a dollar value on it or in uganda a shilling value value on it so we worked out that if we could do this we could trust people we could get them to run these solar panels um and these systems and charge out there but how are we going to pay well there are there's a big story to this but integrating mobile money payments is fine if you're going to go for one mobile network operator but aggregating across lots of them much like the problems that we have in the iot with communication between standards being difficult you have communications with standards being difficult so there are right

Chris Gammell: services you have to do it you can't just like integrate 20 sims you because you have to also maintain that over time and firmware updates and all that other crap you have to do in any kind of network like that right

Stefan Dzisiewski Smith: but also it's the kind of thing where they're like oh you know it's a bit like talking to chip vendors so you're going to do 250 000 transactions this month right now i might do 10 initially because i'm trying it with one person who we're trusting right so trying to innovate in a very difficult space so we then looks at premium sms so the way that people vote for i guess x factor or whatever else or buy ring they used to buy ringtones um they still do a lot in ukanda the problem with premium sms is that it's very low return it's meant to be used or it's generally used to buy and sell goods that don't have physical delivery so the profit margin on them can be very high if it's a digitally reproducible good a marginal cost is

Chris Gammell: low you're saying from one item to the next right

Stefan Dzisiewski Smith: you're quite right yes the marginal cost so once i have the rights to a crazy fog ringtone i can sell it as many times as i like within reason uh paying the license fee or whatever as i go so the fact that i only get 50 perhaps return and if you're talking about aggregating across all network providers in uganda last time i looked at a quote on that um the return was 30 and you got paid six months in arrears oh wow and that was a genuine quote and the best quote that we got so yeah there have been challenges

Chris Gammell: cash flow issues yeah i mean there have been

Stefan Dzisiewski Smith: challenges and that's something that the team has now worked on and integrated much better we spent a lot of the time actually mocking it up by getting people to send multiple normal texts so we weren't actually looking for revenue we just had to make it cost them the right amount to adjust the market elasticity point to a point where it was representative of the kind of demand that we might actually experience yeah um so that was amazing i i joined them i did a couple how many trips did we do i did one trip to uganda three weeks out there um soldering and having soldering interrupted by the generator going out um spending my evenings gambling with other people who i could find who would willing were willing to bet with me on which lizard would catch particular flies in the hotel lobby nice also staying in a hotel where the hotel wi-fi um uh what was the name of the hotel it will return to me are the mount elgon hotel in amale which i can highly recommend if you ever find yourself in amale it is the number two hotel in amale i think on trip advisor uh out of two hotels that are on trip advisor there are many smaller establishments right but uh it's it's it will always be my favorite because the staff there were just so very accommodating but they had a mobile internet there so again we're talking about a place where the proliferation of telephone lines is very minimal right pretty much all of the internet as i understand it goes through actually one of the hotels in entebbe so the main fiber feed i think this is what one of the the telcom engineers was telling me i think it goes through the hilton or something i i might be wrong on the details but it's not widespread and this is another reason why mobile telephones are very um

Chris Gammell: yeah yeah it's the only infrastructure right it's that is that is what it is i mean like yeah that's actually why when you read about mobile speeds in a lot of countries it almost always outpaces the the wired stuff because the it's much easier to drop a tower not

Stefan Dzisiewski Smith: than dropping that fiber right yeah you drop a tower nobody's going to steal you only have to have the tower you don't have to run lots of copper that might be valuable or that people might try and hook on to or right whatever else and you just have to pay for diesel or kerosene to run the the generator um and for somebody to guard it and that's a relatively cheap thing to do right yep but so this this hotel we're staying at and we're like okay what's the wi-fi i mean the wi-fi was abysmal and the reason it was abysmal is because it was uh a 3g dongle and a router yep and that was it yep in the hotel lobby and sometimes you could get a bit of wi-fi in the rooms if it was the right phase of the moon or whatever else uh and as dave would say if your tongue was at the right angle right and so every couple of days um we would try and log on to the internet and it would just come up with the orange uganda uh splash screen saying please top up and you're like oh man this is uh can

Chris Gammell: we sorry could we put a bit of money in the meter over here is there a satellite internet i need a data

Stefan Dzisiewski Smith: sheet wow i mean data sheets were out of it i it was very small amounts of email and making sure that we were well prepared we had a full medical kit and everything yeah yeah because you just need to be ready and you could be a good drive from a hospital so if you get bitten yeah you need to you need to be ready but yeah and in the end we actually so we started chatting with the hotel staff i mean we were there for three weeks we saw the hasidic jews who were there to do missionary work come and go we saw the american field doctors who were doing some amazing cataract surgery out in the bush there come and go but we got friendly with the hotel staff and they showed us the router and it had a connection for a an antenna something like a reverse sma or something like that or smb and we had a an aerial with us and we plugged in the aerial and we were oh we were flavor of the month suddenly you could get much better internet and everything else and i think they were they were rather upset when we um when we took it with us well we took it with us into the field i have a feeling that we left the actual antenna behind when we went back to the uk yeah um but it was an amazing adventure and india was equally uh simultaneously frightening and absolutely marvelous i mean i remember sitting in a in a market sort of area of town which part of india absolutely true so this sorry uh this is in bangalore or bengaluru as it now is uh just outside of cricket season which was infuriating for me because i really wanted to go and see the royal challengers bangalore in the ipo yeah they got some good good cricket out there yeah they are they they are a a nation of cricketers um but we were in the uh i think it was the end just off the mg road in in which is mahatma gandhi road mg road in bangalore there's this kind of big market area we can buy anything i mean we were buying uh uh micrest d cards at very very unusual prices and with really sort of like that you could buy anything you just had to go down some very dark and frightening alleys through sort of open sewerage and all sorts of hard pass hard pass yeah dude it was it was it was real um but also really really interesting to see that these are just dudes and ladies getting on with their life right this is yeah yeah you know it's just like everybody else and so we were there we needed to get so we got these units out into india uh and we were doing integration there and that handles on the units were snapping because of the load they just it wasn't quite correctly made so we thought oh my god what are we going to do we're out in india it's not like we're going to be able to go and get some more precision plastic parts done so we just drew up a couple of plates of steel that we wanted cut and went to the market and it took i would say three or four hours that might be an exaggeration but it felt like three or four hours sat about waiting for people to talk about steel and to cut some steel and we ended up getting the steel cut and we had to take it to another person to get it and sat down a dead like a blind alley with my friend damon uh gently worrying about whether my life was in danger damon's traveled more significant more more widely than i have and was quite calm at least on the outside um but we were sat there waiting to get these things drilled and then in the middle of the training the call to prayer uh came uh um and we were in uh sorry well it was hang on a minute no it would have been well it was it was the muslim uh guy who we were working with had to go to temple uh which temple i can't remember because obviously you have very many for the various but he had to go and so he said to us can you watch the shop and i said well i uh i guess we have a choice you're sort of halfway through a job for us is anybody gonna come not really kids sort of hung around and i just yeah we sat in this alley waiting uh for the dude to come back and he did and he finished the job did a great job but we bolted these these reinforcements onto the units and i imagine that some of them are out there uh all set up like that it was you know great experiences but uh yeah something that i've now again moved on from because we got that to the point where we got it to the point where we had working uh proof of concept but it yeah the the crunch was really boots on the ground

Chris Gammell: so we needed people integration testing all that stuff right and and to be running it day to day

Stefan Dzisiewski Smith: even beyond integration testing how do you since the business model that we had would involve an awful lot of agents and relative and basically having a business that was largely in either india uganda or wherever you're deploying the units not in london it's not about having lots of you know middle class european white dudes sitting about talking about how wonderful they are it's about people providing a service where it's required so you actually don't need middle class european white dudes but what you did need was to come in and take the business development not in terms of going out and you know talking to vcs but the business development in terms of making the business work in the countries we were working in to come along and so i said look i really think that that's where we should be putting my wage money um we should not employ me and we should have somebody come in and do that and there was a little bit of back and forth um it wasn't a universally popular decision at the time but it wasn't ever it wasn't like i was done with it or like i was unhappy with it or i wanted you weren't walking away right it was about kind of the cause the company you know um flourish and so i i left and the whole time i said i've been doing sort of freelance work very similar to the stuff that i'd done at jason brew studio but the whole while i'm doing that i've been doing bits and and specifically bits for bare conductive um which is a company who do uh capacitive sensing and uh otherwise dull and dumb surfaces into interesting and uh intelligent services really quite right but

Chris Gammell: before we before we get into that i have one or two more questions about the solar thing oh yeah sure um so so so what was the level of integration i mean were you doing like the mpp mptt type stuff as well or was it kind of off the shelf integration type stuff so with the initial units we were actually

Stefan Dzisiewski Smith: using uh mppt dev board from nxp uh for a which was uh an arm cortex m3 i think at the time which shows you how long ago it was because nobody buys and sells m3s anymore but uh that chip actually got discontinued before we could integrate it into anything um and the main part of the the design that needed work was the power distribution with so we were using usb ports for power distribution um as a sort of uh as something approaching a standard although we're not talking about plugging in it both android and a phone here we're talking about strange uh i don't imagine you've ever seen a saga telephone but they're very popular uh in aganda or some other promotional phones that are given away or produced at low cost in india so you have to be able to plug into a wide variety of feature phone sockets and we had these bizarre we called them octopus cables that sort of had every possible

Chris Gammell: connector on the end oh but but at the end of the day it was mostly five volts of various current

Stefan Dzisiewski Smith: draws and then the weird ending connectors yeah exactly the only thing that uh doesn't the very old nokia's um don't do the the five volt but we ended up i mean you also have to shoot a little bit for

Chris Gammell: the future and they are less common the eu standardization kind of really pushed that stuff along which i'm grateful for because yeah that really changed a lot of things for mobiles yeah it's

Stefan Dzisiewski Smith: true but yeah so um a lot of the custom work that i did was so there was some interconnect stuff in there there was power distribution and management um and switching and then uh having a central so we actually used a beagle bone black initially although this is all stuff that's getting more integrated now and damon did a lot of the integration to allow us to hook up to the cellular network and integrate with various payment providers and do tracking and in the field um updates and logging to get data back so that we could be back in london saying you know to potential investors look this is actually happening we're getting a hundred charges a day in some of these locations that's profit for our you know that's a good business for our agent and that's profit for us because we take whatever percentage of the income uh but the mppt we actually ended up using uh uh i suppose i can plug them on there but uh rock solid solar who are uh an australian um uh solar never heard of mppt company i don't know maybe dave has i can't remember quite from where they are but we used a module from them for a while and now it's got to the point where the proof of concepts are out there and the business is being developed and the volumes of units are no longer in the tens or so on because also it's about failing fast so yeah we weren't particularly bothered i'm sure i can do the mppt but proving that i was a mppt isn't worth the time in the field finding out how people use it you know with my studying at the rca wasn't just posing or trying to get into parties or getting good connections for later in life it's about understanding the way that you're always selling a relationship even when i was working at high voltage place we weren't selling high voltage power supplies we were selling the relationship that you can believe in us and we will deliver eight ppm per hour drift right which is crazy but on a high voltage power supply but we will deliver it and we will deliver you know the switching speeds that you need for your mass

Chris Gammell: spectrometer because you trust it right so it's about it sounds like you were doing you know the way you were talking about the problem too i mean you were looking at it holistically you weren't necessarily just approaching i mean you had the e background but you were looking at you know actual field tests actual like understanding that the bike wouldn't quite work trying to find these other methods like that's a holistic product design versus just just the one piece yeah and phil the guy

Stefan Dzisiewski Smith: whose idea uh the company was the bike was his idea it was his pet idea and it's a beautiful idea but we had to explain to him that we didn't think it was long-term viable and that's not an easy thing to take somebody who's brought you on board been very generous in employing you um and said can you make my idea work and turning around to them after a relatively short amount of time and saying not we don't like your idea but we don't think your idea is best for the company that's a difficult initial conversation to have i mean phil was incredibly understanding about the whole thing

Chris Gammell: but it does yeah yeah well i know someone that worked at a at a not not quite the same kind of it was you know a battery charging type of scenario but i had a friend who two friends who worked there and they didn't they didn't quite do that due diligence they weren't part of the due diligence at the beginning they were just implementing and they did the best they could but at the end of the day it's like they you know it was like physically impossible to deliver that stuff and that that's where then and it basically ended up affecting the company in the end that it just it couldn't deliver on what they said they could so all the investment money that went into it

Stefan Dzisiewski Smith: it was wasted well this is the thing the road to hell is strewn with good intentions and you do sometimes have to be you have to know what is it kenny rogers would say you have to know when to hold them know when to fold them know when to walk away oh my goodness how can you possibly say

Chris Gammell: you're from the uk if you you know there's such a critical part of uh you know u.s culture kenny rogers

Stefan Dzisiewski Smith: was big in the car the guy alias who drove us around uganda who looked after us and fixed stuff for us uh he loved kenny rogers and the christian worship cd he also had a bit of dolly part in there he was muslim yeah but he loved the christian worship songs and he was happy there was no reconciliation to be done here he was like i like them they're good songs so i was like okay that's cool all right well you're driving you get to choose i guess you totally got to choose it was those three cds for three weeks boy howdy so yeah i heard a lot of the gambler by by kenny uh and i still love it which i i guess goes to show that kenny rogers had good things to say there you go

Chris Gammell: yep exactly but yeah okay so so bear conductive so now you're you said you're still consulting and uh what is bear conductive so bear conductive uh we i don't know if i have the company line

Stefan Dzisiewski Smith: i'm never very good at this kind of thing but from my point of view we take passive sensing and use it to generate data in a meaningful way from in from unexpected places so the nice thing about cap sensing is that it's relatively cheap to make relatively big things um yeah and you can then sense things in either a cheap way or in a way that's integrated so that you don't um you don't make people feel like they're being spied upon it's also inherently uh it's inherently anonymous it's not like a camera where you can identify people particularly from it so the company started however as a student project so we're going right back to the rca again which is this terrible cult i am part of and will never escape please help me it's good it sounds like it was networking oh you know it's fantastic but it is you have to be careful of not trapping yourself in an echo chamber um it's very easy to all sit around coming up with ideas and deciding that they're they're a marvelous and good idea but you do again have to be open to people from outside one of the things we work hard at at bear conductive is employing people who haven't been to the rca um to make sure that we we don't develop a monoculture but that's good so it started matt uh johnson um is a good friend of mine from the rca so you've got matt johnson bb nelson isabel lizardi and becky pilditch so those four uh were at the rca same course as me same year and our course has two end of year outcomes at the end of the first um sorry two outcomes in your final year halfway through the year you complete a group project and then you do a solo project to the end and their group project was a non-toxic non uh sorry non-toxic carbon based uh carbon and water based uh electrically conductive paint and the idea originally was to investigate electronics on and near the body now i have to be very clear and say before anybody uh gets the wrong idea the paint is definite that we sell now is definitely not uh approved for skin use and we wouldn't advise that anybody did it um it is non-toxic there is very little likely to i mean i've eaten the paint um and it's not killed me but well i mean it might

Chris Gammell: have turned you a little weirder but you know that's just the baseline was very difficult to detect

Stefan Dzisiewski Smith: before though so it's hard to tell you know it's all about deltas with stuff like that it's all about deltas that's right right but precision versus accuracy well exactly exactly uh i think both are shot on on this particular model but um so they had this idea of this electric conductive paint and they did some really beautiful demonstrations and they came from this conceptual idea of what can they do you know on and near the body it wasn't how do we make the most conductive paint that's going right there are loads of conductive paints out there and inks and carbon nanotube this and silver loaded and you know copper loaded and graphite load all these things right so they specifically

Chris Gammell: weren't trying to do like the conductive pens where you're like hooking things together one point to

Stefan Dzisiewski Smith: another that kind of thing right although they did do sampling circuits that way i think they very quickly realized that its use as a sensor was the most interesting thing about it um right or opened up some of the most interesting applications but the initial projects they started so they got some seed funding and when i moved back up to london to go to jason brew studio um my friend matt was looking for a place to live and we moved into a converted warehouse in manor house in london which is like a crazy warehouse party scene i have to say excuse me i am done with warehouse parties for the rest of my life my god when they start on thursday in the morning and run through to the following wednesday and nobody shuts up like i mean solidly that's not my style we're talking aggressive dubstep drummer bass there was a big summer of of i think they call the mcat uh it was mcat the name of the the drug which was um oh what's the the heroin substitute it's not that so it's not methadone but it's it's it's methadrone methadrone uh was this drug anyway there was loads of that going on that summer it was to call it messy is is beyond belief i did meet my girlfriend whilst living there at one of those parties but we have both since uh moved on and i'm still with her so uh good things can can happen in strange places right right of course so but matt and i moved into this this converted warehouse and there's a huge space and so bare conductive started i would go to work at jason brew studio and they would come to work in my flat and then we would oh weird yeah we would swap uh back at the end of the day and i still remember coming home one day to find that my house was entirely black because they'd put a mixer and had some sort of incident with a lot of carbon black which had gone all over the inside of the flat and i tell you what when i left that flat i was still cleaning it down and it goes everywhere so again who knows what what troubles are science is messy science is messy man science is definitely messy but i you know what you were talking about what their aims were i i think the nice thing is that at the very beginning there was a limited understanding and on their part about the existence of conductive paints and inks they they had vague ideas that you could buy like a silver paint pen from matlin to fix things but not of the breadth of the market or prior motivations and industry sectors that use them right yeah and the um naivety if you like in their understanding of what had gone before but their solid understanding of design thinking and of process and of rigor and of a an experimental approach led them to some really interesting areas so they started making this paint and they got to a point where they made some kits where it was point to point like you say but basically kind of a lithium battery you know a primary lithium cell and uh some of the paint and a flashing lead right and they made that into some beautiful cards so they contextualize it in a way that taught kids or people who are interested in playing with the idea of conductive paint about it another really nice thing is it's super cheap to make so you can screen print it and not worry about the fact that you know uh if you take other say silver loaded paints or inks they're they're not really silkscreen compatible because they're very thin with the solvent right and they're not cheap enough that you can dump you know a tablespoon of it into a screen and flood it so they started to play around with these things and they came to me and they said we've done some things with this arduino and kind of cat sense library what do you think we kind of want to start to to play with it maybe do a version of a board with it maybe that makes audio so you can kind of paint a poster and touch it and then it plays audio and so i started to look at some of the examples and actually i did a lot of the research based on um modules that were available at places like adafruit and spark fun purely because i in my freelancing had gone from so if you remember back when i said i was talking about microchip controllers and and assembly coding for me the power in doing when i was doing development work for clients of using the arduino platform is that for the vast majority of challenges i was being thrown there was enough horsepower there and it meant that i could hand over something that if they then wanted to change in the future i didn't have to be part of that right i could hand over something where they could find huge numbers of other people who could then work with it it was a free tool chain you could get the most professional cc++ embedded programmer in the world or you could get um somebody who'd come through a design route and it gave my clients the ownership so i was like look we should maybe do this with arduino and then if i pick and choose from modules and so on that are out there then we know that there's a reasonable established market you know that the availability of those parts is going to be reasonable and that there will be libraries that people will be able to access to take control of it themselves so if they want to write you know their own bit for it or they want to use somebody else's library for that part they have that choice and that became the touch board which was the first um circuit board which is the yeah that's the

Chris Gammell: some people might have seen right it's got like the it's not castellated edges but it's like uh touch points on the edges that you actually hook wires into yeah it's got 12 edge-plated electrodes along the

Stefan Dzisiewski Smith: top uh it's got an on-board mp3 player it's got an at mega 32e4 so it's like the old uh retired arduino leonardo uh you've got a micro sd card lithium polymer charging circuit on there so you can have a lithium polymer cell you can power it from the cell you can recharge it via usb you've got audio out and you can do things like usb midi it's got on-board midi synthesis actually for me it's it's now the art if i i'm going to drop an arduino into things i often use it uh partially because i know you know that it's reasonably well designed i hope um but also uh having the micro sd card already on there is really useful for anything where you're doing data logging so i actually use it as a logger a lot yep so yeah we kick-started that and i i think at the point where we kick-started it it was the most successful kickstarter in the uk up to that point 112 000 pounds i think i'm getting all the figures wrong no doubt matt will be uh spitting feathers when i see into it well when he hears this but i think it was something like that okay so it's just it's just it's just numbers man yeah well it doesn't make it look i didn't join the company in order to i i joined the company because when i was freelancing and when i moved on from buffalo they were just my favorite freelance client and they were friends and they i it's exciting stuff

Chris Gammell: and probably good uh good users of the product stuff like that i i find it really interesting so

Stefan Dzisiewski Smith: we made this thing largely for makers and got it out there through um through kickstarter but we knew that we had in understanding the capacitive sensing side of things so i actually ended up writing a whole new library for the npr 121 which is the cap sense chip on there where i really stripped down and tried to really understand what was going on not just a parallel plate capacitor thing but how do you optimize sensing how do you give people the kind of apple out of the box experience whilst leaving the lid open enough that people can go in and get right into registers if they want to and i wrote it yeah for what we wanted to do and in fact off the kickstarter we got contacted by a number of um industrial potential industrial clients saying whoa you can sense of distance with paper and we're like yeah i kind of hadn't thought about it that way but what we had shown them in our kickstarter video was very cheap sensors right screen printed paper with with cheap paint electric paint um and a board that we were selling for whatever it was on the kickstarter and we'd shown it sensing over distance and sensing through tables and sensing through glass and i think that the engagement that we got there brought us on and brought us into the interest of a number of industrial clients and that's where we're now doing a lot more of our r&d so we are bringing out a so we're bringing out another consumer uh face product in i think october sorry september maybe early september i think um which is very similar to the touchboard but for the raspberry pi uh and it's specifically designed around the pi zero so one of the things a lot of our customers want to do is to have polyphonic mp3 playback but this also more importantly gives um us a bridge between the capacitive sensing um and things like well an operating system fundamentally so the internet

Chris Gammell: all of that kind of space yeah exactly but also then a wider variety of languages so you're not

Stefan Dzisiewski Smith: stuck with arduino so we have um cc++ library we have a python library we have a node library in the works uh so some javascript which i know you're a fan of well yeah of course but i mean i mean

Chris Gammell: like honestly these days for a lot of this stuff it's it's especially if you're you you guys are basically bringing people in into the you guys are recruiting more potential than amp power listeners and that's and that's fine with me honestly like it's it's bringing people in that are software and want to get hands on the real world so it actually i think it makes a lot of sense for uh for this kind of stuff yeah and the question about that though is how do you get a raspberry pi zero because those

Stefan Dzisiewski Smith: are kind of uh purple screws right there is a i like the expression purple screws i always say unicorn poop or hen's teeth uh but purple screws yeah i've heard hen's teeth too yeah um unicorn poop i might have to use that yeah it's it's it's tough to find um not if you watch that squatty potty video have you seen that one i have not uh that sounds that sounds potentially horrendous is it going

Chris Gammell: to change how i feel about things i don't know um no it's actually it's it's uh it's very catchy as surprising as all that that sentence was it was it was writing down the words squatty and potty

Stefan Dzisiewski Smith: next to each other i didn't expect to do that today they're good words they go well together you know a little bit uh a little bit of surrealism in your life why not you'll see it nonsense now and then is relished by the wisest men uh that's what they tell us um i don't know

Chris Gammell: where we got before that uh uh we were talking about the raspberry pi zero and why you can't get

Stefan Dzisiewski Smith: ah yes well we were very lucky so we have good contacts with um the people of the pi foundation

Chris Gammell: and raspberry pi trading ah it's a british british thing again come on man we we they came to us

Stefan Dzisiewski Smith: when they were getting ready to release the pi zero and we talked about doing something we've actually held off releasing a little until the audio on the pi improved um which is i think it's actually now current in the latest version of raspbian but it's certainly something that we enable um so we get 15 and a half bit audio out now which is a lot better than the onboard audio used to be uh which is let's be clear none of my work some great work going in by the hard beaver uh people beavering away uh at pi van or working on on the raspberry pi linux side of things yeah but yeah they give us a bit of a heads up and some boards and also i just keep an eye um again i don't want to plug too many companies but there are um there is a website where is my pi zero which aggregates where your pi zero may or may not be um in terms of who has stock uh but places like jamie madd at the pi hut i've bought one of him last week it's a case of of tracking it and i mean it is one at a time um i think it might be tougher in the states because i think you've only got adafruit reselling

Chris Gammell: over there right uh i've heard that you can get them at microcenter that's what someone told me that might be right which is a great company i love them i yeah i do too they're there they are basically the they've turned themselves into like quasi radio shack without the components like at least for dev boards and stuff you can get a lot of stuff there which is really nice i i was lucky i think

Stefan Dzisiewski Smith: it was the the head honcho at um microcenter i met him at maker fair new york a couple years ago and he was just a straight up brilliant dude uh really really nice uh and he had a lady who worked with them who i think she'd started working in the shop and was now she worked her way up quite high in the company i just seemed like really reasonable people i know there's a lot of people who are wonderfully reasonable in the kind of maker community but uh i was really struck uh by the microcenter people really nice um but yeah so that's how we got our hands on the raspberry pi zero but this this add-on will work with other raspberry pi as well but that's kind of that's one area so we are doing that but we're doing some fairly serious r&d now so i really i enjoy kind of wrestling with physics and trying to beat it into submission and i'm getting the opportunity to do that a lot i'm doing a lot now of straight up field design so it goes right back to stuff i learned doing the uh the high voltage stuff things like how to do field confinement how do you make one-sided capacitive sensors how do you um take make capacitive sensing work over a long distance you know if you're going to actually have rather than putting localized stuff and having digital data how do you uh confine the field over coax and things like this without losing sensitivity or loading the lines too greatly how do you make all of that compatible uh with emc regulations why is any of this important what do we really want to sense and we are working sadly with companies that i cannot mention but i'm looking forward to in the next couple of years actually releasing products with these companies and saying look this is what we're doing at one end of things but we're still releasing these open source um uh and open so open source software and open source hardware products for the maker community as well because we get such interesting stuff back off them i mean another great thing one of my favorite things uh with the paint is it allows you to iterate and create custom custom interfaces really quickly so if you want to just say oh if i lay out buttons like this does that work you can screen print it put it on the touchboard blob it in you can be done in 30 minutes right and that is a straight up ux prototype we get people doing things like um creating custom musical instruments for people with profound physical disabilities and you just think whoa i am so glad like it's totally worth putting the effort in and making that board and supporting it if it gives people the opportunity to do that we get people doing beautiful artworks we get the people um making lovely posters with their kids uh you know just spending a weekend drawing a beautiful poster out in the paint and then connecting it up and putting i don't know farmyard sounds on we get kids doing yeah school projects with it um

Chris Gammell: yeah it's just like automate not automating the world it's it's kind of livening up 2d things right

Stefan Dzisiewski Smith: yeah exactly and you can do it in three dimensions as well i mean that's where we're starting to do a lot of this r&d um is on three-dimensional surfaces and can you know field containment within volumes so looking at stuff passing through volumes and it's great it's great fun a spherical cow uh

Chris Gammell: extruding milk in all directions or whatever that is i don't know that sounds like a another

Stefan Dzisiewski Smith: crazy does that go with the squatty potty that's i don't know no that's a physics joke somewhere i

Chris Gammell: forget what it is yeah it's yeah it's something about a that's pure it's like how everything in e&m in physics e&m kind of comes back to like you know a infinite plane or a pure sphere floating in

Stefan Dzisiewski Smith: space type of thing no it's true it's kind of like i mean picasso really took that to town didn't he the cubism type thing you know a reductionism to primaries uh primitives so so so this chip that

Chris Gammell: you're using the mpr 121 i've got i've got the the um the data sheet up and this is a hefty set of registries i mean it's 128 registries yeah this is pretty yeah that's crazy 16-bit register is that

Stefan Dzisiewski Smith: right uh yeah so you get 10 bit 8 bit get 10 bit read back uh i think 10 bit 10 bit yeah two three okay oh yeah okay um but and it's got some incredible filtering on it it's a really interesting chip actually so it's originally um a freescale part um and then there was the whole nxp and freescale thing happening and in amongst that it got completely jettisoned from the company so i had assurances from freescale about the longevity um and they jettisoned it and we went oh man you know i'm sure we could obviously there are a lot of last time by last time oh man rochester electronics not my favorite company in the world um but we looked at that and we said okay look we can we can do things differently we could take the whole touch board down and we could do it all with integrated um uh touch peripherals right that that you know now with the arm yeah a lot of the micros that have them well they're trying to differentiate between the same core amongst lots of different manual i mean soon there'll be no need to differentiate because after analog bought linear today they'll they'll only be like they'll be the monolith chip company and that will be it you had to bring it up didn't you i'm

Chris Gammell: sorry well i do pay a little attention and i i was i was i was waiting to bitch about that until next week but sure yes we can yeah that's we should acknowledge the inevitable my favorite my favorite chip company has been absorbed by one of my less i don't dislike analog but i didn't like them as much

Stefan Dzisiewski Smith: linear are really beautifully designed i i would love to meet the person who does their capacitive sensing um dev boards they have really there's a really nice pro uh move on one of them that i'm working with at the minute where you have the chip that you're working with and you have you know i think it's an lpc processor on the side and usb connectors and ground planes and it's all properly done and the ground wires are stitched appropriately and it's it's well designed it's got square pads on the top of the breakout area so that you can do surface mount and circular holes underneath so that you don't bridge you know it's been well thought through yeah and it's got a little like yes well sure it's got a little bit on the side where they've got two chips that's just held on with pads in opposite corners and those are your those two chips are your samples so they're just tacked down and then you can just easily pop them off with a heat gun so you know where they are when you need them and that

Chris Gammell: was a really nice touch i thought oh that is that is a nice touch okay i'll give you that it might not be enough to to sway you and we'll have to see what happens with analog and linear i've used a lot of analog parts it's just that i just don't i didn't like them as much as linear and linear uh you know yeah they've got their quirks too so well i really like linear data sheets um i have to say they're

Stefan Dzisiewski Smith: some of the best uh but lt spice if they touch lt spice oh man well i think that the world might end if they touch lt spice i i yes i've used it for several things well we've had mike on the show

Chris Gammell: mike englehart uh who basically is the maintainer oh really main trainer i missed that somehow crater too oh it's it's a that if people want to listen that was a fantastic show and you know to analog's credit we had uh hank zumblin on uh hank's great uh he's one of their their lead analog's lead um application engineers um but yeah well let's hope that the the cream rises to the crop and not not the other way around um i'm sure i'm sure i'm sure that'll be fine i hope yeah i hope so

Stefan Dzisiewski Smith: too but yes so we we had all these assurances and they jettisoned it and they said oh this company resurgent electronics will be taken on i was like oh resurgent rochester both start with an r the price is gonna go up 10 times and they're just gonna do end of life stuff with it and yeah look we could go another way and we could use other uh integrated touch peripherals but i'll be honest i've trialed a lot of capsense chips out there if you want to make robust uh interfaces that work no matter what you clip them to and how badly you abuse them and can do really nice proximity and yet really robust touch and they just work and the filtering is all just as you want it to be the npr one-to-one's a great i see but i got in touch with uh the so uh resurgent are really two guys i think with a lot of experience in silicon and duke adapa they're like a ceo um i got in touch with him he's like look we've got no uh need to put up the prices we've got no wish to do so we want to if we can we'd like to grow the market for this chip but we don't need to sell that many to make it a viable product for us we bought it up we've got all of the tooling will be exactly the same we're not changing anything except we're going to mark the top of the chip differently but it's exactly the same thing we're just going to run it through through foundries and they straight up just are they going through distribution or yes it looks like they've got arrow it's just arrow yeah so if you see it anywhere else uh that's not going to be a reliable long-term source and as long as you buy

Chris Gammell: a reel at a time you're fine uh right which is what you're doing anyways i'm sure yeah we normally

Stefan Dzisiewski Smith: have a few reels in stock and and we just do it that way and i'm sure that everybody else who's on the breakout board side of things i'm sure adafruit and sparkfun are doing the same thing and keeping it going and i think it's a really nice chip for people to learn cap sensing with because it just yeah by the time you're questioning the real details of what you can get out of it you now understand enough that you'll be able to look at other options and make choices yourself but for me as an external peripheral

Chris Gammell: yeah i think it's great you know i've actually never thought of this before but this is almost like the silver lining of the the uh of the foundry model right where it's effectively so you said these guys these two guys have uh fabulous experience which is great but effectively it becomes like a uh a what's it called capital it's a venture capital move right where you buy the ip and you can keep making it because it's on the same process whatever and the silver the true silver lining is that there's almost guaranteed in that situation to not be any changes to the underlying silicon because why would they change something it's just a cash cow for them exactly so us as the users it's like all right as long as the process is still available at some foundry and it's qualified against the other chips then you almost could have an unlimited supply which is of course big asterisks on that but

Stefan Dzisiewski Smith: well i think they promised me until 2022 i think oh no 2017 no hang on sorry go back 2022 next year yeah 2017 would be a worry sorry i'm adding and subtracting five in my head no they said until 2022 and potentially well beyond so for us that's great and that's probably yeah that's probably based

Chris Gammell: on foundry kit changes right where that's probably what the assurance they got from whatever their

Stefan Dzisiewski Smith: yeah in terms of process of whatever where they're currently yeah manufacturing i mean i don't know too much about ic design that's for sure but um or manufacture but uh yeah they and they seem again like just people who spotted an opportunity and are making it work so uh yeah it's all good for us

Chris Gammell: this is yeah this is actually this is very interesting i'll have to look at this so

Stefan Dzisiewski Smith: resurgence semi.com yeah that's it i think they only pretty much have the npr 121 and maybe the

Chris Gammell: npr 03x which is the the lower electrode count version of the ic i'm sure maybe there's maybe there's just a general business for this kind of thing too it's just kind of being like the shepherds the the warden of the chips yeah as the i think anything that stops it being rochester electronics

Stefan Dzisiewski Smith: sitting on a massive pile and then selling them off at four times i mean i i understand their business model and i'm sure that they uh employ a lot of good hard-working people but it it really

Chris Gammell: hits me you know you know what we're not we are not the customers for that it is the military is the customer for that and aerospace and that is so it's fine if that model extends to industrial and maybe commercial like this which is seems like it might but yeah the i can never i will never and can never pay aerospace or uh military prices so neither can i and it seems unlikely with the decisions

Stefan Dzisiewski Smith: i've made in my life that i will ever be wealthy enough to make those kinds of decisions so

Chris Gammell: it's i think that'd just be a irresponsible product decision too right why would you why would you do that it's better it's your time's better spent redesigning which i'm sure you're already doing

Stefan Dzisiewski Smith: with with these things well we're working with other ic's right now we're happy with the b2c stuff um we're happy using that and look if we get to a point with a b2b project where uh where the volumes are enough that'll be enough to to well not guarantee it indefinitely but certainly to give us enough of the market for um resurgent that we can be a big sort of customer for them and and then they don't have to worry and we don't have to worry and if they're right we're not in a position to do that i wouldn't be uh an engineer worth paying if there wasn't another way i could achieve similar

Chris Gammell: kinds of things right yeah any thought to go design your own uh discrete circuits to do this kind of thing or i mean obviously it's not worth it but have you have you gone down that that path at all

Stefan Dzisiewski Smith: uh not so much i mean there are lots of because of what you can do with uh the integrated peripherals in in a lot of the the microcontrollers there's not a lot of gain to be got out of out of doing it discreetly um i mean you can it's a lot of the principles actually are are very similar to the kind of ramp generators within um within uh things like switch mode power supplies it's just different frequencies and then your your ramp is generally your output current fed ramp and then you're doing bursting on that and and uh acquisition through sort of integration um but it's not something yeah i don't i can't see there being a motivation for me doing it i do know that there are companies that have done their own uh there's a flooring company based out of uh somewhere in finland i think that was based on a university project that used i mean dear lord their bomb i really hope that they're using something else in the final implementation there was like a a 20 dollar analog multiplier in in the design in the um in the academic paper i was like whoa you were way richer at university than i was a 20 dollar multiplier like someone got a research grant and how right and it had custom magnetics in the output stage to do impedance lowering i think i it was it was just like a you know when you go to a firework show and they have one of every type of fireworks you get a feeling towards the end they're just trying to cram in another different type of explosion i feel like yeah that might be what was going on on the board maybe somebody was proving that they really really really deserved a phd or something but uh or maybe maybe it grew that way i don't know i mean you can't you can't say sometimes having a circuit where you can actually probe a lot of the intermediary signals and you can adjust almost every aspect that you want gets you to a result and then you go back and you think right now i know what the result is and i know what the kind of end game is yeah i can strip out a load of this stuff that's redundant but you know it's a bit like when you you build any kind of test rig or development rig you don't build it with an eye on mass manufacturing when you're making it you're leaving test points around your you know the amount of times if you think about um with the level of integration that you've got now that you just beg that you could get an intermediary signal out of a highly integrated ic uh and sometimes you can get them on debug pins uh things like that but yeah i guess if you're building your own circuit you you would be silly not to leave access to the

Chris Gammell: signals that you want to see so right right that makes sense right yeah kind of kind of like internal probing by making everything external huh yeah exactly exactly the old-fashioned way yeah but yeah well uh we are uh an hour 45 in maybe we should uh i i i dare ask is is there other

Stefan Dzisiewski Smith: stuff are you doing other things man um i am doing a few other things but mostly it's just um sort of side projects i like helping people out so if people email me i very rarely don't mike's always been hugely generous with his time and i try and pay that forward to other people i do little bits of freelance here and there just stuff out of interest i did some hacking of a i wrote a library to interface with an induction cooker uh earlier in the year for somebody who wanted to make their own design front end for it that was an interesting challenge working with really horrible i squared c but it's um like i squared c that was so far out of spec that it was i had to write my own uh kind of like an interpreter interpreter yeah so i was using um i think it was uh an open logic sniffer and then just dumping the data out of that and then running it through my own interpreter if i tell you that the clock idled low in this particular setup it was it was super weird and and very clever i mean it was allowing the the chinese designed front end board to do button pressing lead driving through charlie plexing and i squared c communication all out of a very small number of of pins on a cheap eight bit micro so props to that i see but it was dirty as all hell no i i think we better let the listeners if you have listened this far i am very impressed uh to me just rambling on in my in my in my bedroom talking to you here uh but yeah i that's good it's this is

Chris Gammell: this has been this has been a wild ride of all the way from uh high voltage up to capacitive sensing to art art panic attacks i think oh man dear lord don't don't yeah i i may

Stefan Dzisiewski Smith: have to have a sit down i'll have a sit down with a beer and try and forget it all in a bit i was

Chris Gammell: gonna say yeah you go go play some darts man it'll yeah it'll chill you out just and say hello to the

Stefan Dzisiewski Smith: queen on no i don't say hello to the queen i did once say hello to boris johnson on my way to work he was cycling morning boris i said and he just went and cycled off yeah he rode by he rode by the

Chris Gammell: meetup when when we did the amp hour meetup someone's like hey look at that yeah look at that out the window boris johnson's riding his cycling by and if i would have known now i would have like just shot him something yeah yeah yeah he's a horrendous buffoon let's not get into politics that'll take us all night no of course of course but uh yeah but it's been great yeah well thanks for coming and uh we'd love to have you back sometime maybe maybe once dave's on the mend we can we can talk more about cap sense and anytime and especially i'd love to yeah it's been exciting hearing about the

Stefan Dzisiewski Smith: the future stuff too it's great chat to you it was great to meet you uh in belgrade i think the nice thing about the empire is is always the variety of of what's on here um and it's nice to be able to contribute into that and hopefully yeah i hope dave's on the mend uh and i hope he is

Chris Gammell: listening uh he should be he's got he won't be he won't be if dave if dave's listening right now i couldn't i can't even say that because if i say anything too weird people would be like dave did you hear blah blah blah on the empire and so it's fine well i hope he's getting well anyway yeah

Stefan Dzisiewski Smith: uh where can people find you online uh so i guess bearconductive.com is where i do uh is where i do my day job uh on twitter i'm stefan dz s-t-e-f-a-n-d-z uh and i do have a website but it's so horribly out of date um that it's it's not got much but it's not that bad uh it's so that's b-y-c-g-w-t-s-f.com which is the worst domain name i ever bought which stands for bet you can't guess what this stands for dot com i i regret i regret the the purchase uh but i have never come up with a better idea for

Chris Gammell: a domain name so there you go that's more for me well it maybe it's easier to spell than your name

Stefan Dzisiewski Smith: too maybe that's maybe that's the reason maybe that's it maybe i'm drawn to unpronounceable strings of consonants that's that's my kind of thing yeah well thanks for being on and we uh we

Chris Gammell: really appreciate it and uh we'll talk to you soon yep bye you

Speaker ?: you